Designing and Prototyping

Copyright 2015, 2017 by Collaborative Authors, All rights reserved 

This book is the work of a collaborative group.

Contributors 
Larry Ball (Primary Author)
David Troness
Petr Krupansky
Krishnamurthy Vaidyanathan

Editors
Erika Ball
David Troness
Paul Dwyer
Robert Lang  

Illustrators
Larry Ball

Other Authors, Theoreticians, Practitioners Whose Writings or Teachings have Impacted This Work
Genrich Altshuller
Ellen Domb
Roni Horowitz
John Terninko
Alla Zusman
Boris Zlotin
Lev Shulyak
Yuri Salamatov
Victor Fey
Eugene Rivin
Darrell Mann
Sergei Ikovenko
Simon Litvin
Peter Ulan
Lane Desborough
Clayton Christensen
Renee Mauborgne
Kim Chan

Much of the material for this book was inspired by the thought leaders referenced.  The original intent was to codify the insights of these thought leaders, but the exercise of codification ultimately led to the synthesis of other experimental processes.  This is because codification required recognizing patterns of similarity of tools.  Once this was achieved, the various tools were grouped with key decisions.  Decisions require and create information which flows to the next decisions.  Patterns and gaps became visible during this formative process.  Experimental methods were inserted into the gaps.   The proof of these experimental methods is whether they actually help the reader to identify product or process characteristics that will delight the market.

Prerequisite Knowledge:
Throughout this book, we will be working with the concept of “Function” and “Job”, so it is important to understand the definition of these two terms as they are used in this book. If you have not read TRIZ Power Tools-- Working with Functions please do this before reading this book.

Each of the books in the TRIZ Power Tools book series are designed to be used as an algorithm. 

The Output of this Book is Prototypes and Manufacturing Knowledge
The output of this book is prototypes that we can show to the customer and protection which will allow us to do so.  Once we have these prototypes, we can circle back to Creating Offerings and continue the iterative process of interviewing representatives of the market segment and getting feedback on the prototypes in order to get the features right. 

We should again emphasize the point that to truly understand the market and the features that they want, it is necessary to enter a learning process rather than an execution process.  It is unlikely that we will understand the customer needs the first time around.  The consumer generally can tell you when you have it right, but it is our responsibility to place different concepts before them to help them fully understand their needs.

This problem is equally acute on the business side.  It is important to understand the needs of the business and where they are willing to play in order that constraints on the market segment do not end up in the business and its ability to provide and deliver the product or service in the way that the target market segment would like.  Thus, there is the added duty to investigate production methods as part of the prototyping process.  When we get back to the business managers with our prototypes, they will want to understand their recurring and non-recurring costs of manufacturing.  Will they need to purchase new equipment or can the existing equipment be modified?

Method

Step 1:  Begin with the tools that you already have.

Step 2:  Add to these tools as opportunities present themselves.

Step 3:  Learn how to use these tools

Step 4:  Following hand tools, the next round of tools can be bought relatively inexpensively such as grinders, welders, power saws, and power screwdrivers

Step 5:  Add more expensive tools such as lathes and mills and stereo-lithography equipment.

Step 6:  Install bins which can be used to store the materials.  Get several sizes, allowing you to stockpile a variety of sizes and types of parts.

Step 7:  Make it a common practice to disassemble products and stockpile the parts.

--This allows you to see how other products are manufactured

--This is a great source for odd parts like switches, motors and fans

Step 8:  Store the parts in a manner that is very accessible

Step 9:  Use the parts.  You will best learn where they are and how to use them if they are in constant use.

Step 10:  Become aware of local businesses that sell raw prototyping materials. 

Step 11:  Buy, store and learn how to use the following materials

--Torn down parts from other products

--Fiberglass resin and Bondo™ (Available from auto parts stores)

--Plywood and scrap wood

--Aluminum Sheet metal

--Lexan™ Plastic (now available at hardware stores)

--Plastic and metal tubing

--Aluminum stock metal blocks, rods and sheets (Aluminum is a wonderful material to work with since it can be easily worked with many hand tools)

Step 12:  The local hardware store carries many items that can be used to create prototypes.  Thus it is not necessary to keep a full supply of all possible prototyping items on hand.

Explanation

If you have not done this already, you can begin to build your prototyping tools and materials.   It is not necessary to have all tools and materials at once, but instead, organize as you go. As you become more organized, you will find this to be a big time saver.  

Tools
Many people believe that a lot of expensive equipment is required. A typical garage setting with hand tools is usually sufficient.  This is because prototyping can often be performed using existing sub-systems that require modification.  Having every convenient tool is not necessary in the beginning.  Simple hand tools are often all that is necessary to begin with.  This is because many types of raw materials come from parts that are already manufactured and only need slight modification.  As time goes on, you will want better tools.  Ultimately, they allow you to build better parts and allow you to extend into Aesthetic Prototypes if necessary.  

Materials
In this modern age, there are many raw materials that allow for rapid prototyping and are quite easy to use.  The budding product developer should take every opportunity to learn about these materials.  Make it a point to investigate new materials each time that you go to hardware or other specialty stores.  Talk to others about the materials that they use and the best ways to use them. 

The building of prototypes can be frustrating or satisfying depending upon the resources which the inventor has at hand.  While it is good to have raw material available in the form of rods, tubes and blocks, there are many types of pre-used objects that can be modified to form prototype objects.  Excellent accessibility of these materials will also enhance the experience.

Raw materials can come from new or used products which are designed for completely different applications.  One of the author’s hobbies is gardening.  One might say that successful gardening in the Arizona desert requires innovation as the raw materials are not optimum and the margin for error is low.  Many types of gardening equipment that would work in other locations are completely unsuitable for the Arizona desert.  It is often necessary to highly modify sprinkler components, for example.  When shopping for the pieces, one part may be found in the plumbing department and another in the hardware department as chair leg protectors.  One of the author’s favorite places to look are second-hand stores.  With a little practice, you will see the potential in other people’s cast-offs.

One helpful hint is to create a stockpile of parts, which can be readily modified to build prototypes.  Keep bins of raw materials such as small, medium and large plastic, metal and wood parts.  These storage bins need not take over the workshop and they can be stocked by taking apart unused or inoperative equipment from the home or second-hand stores.

As a precursor to this step, it is handy to have a number of objects at hand that can be used to perform experiments on basic physical phenomena.  It is good to be constantly collecting objects associated with unusual physical phenomena. Make it a practice to disassemble second-hand goods to create useful parts.  You will learn a lot about different manufacturing methods and collect inexpensive parts at the same time.

Method

Recall the offering requirements that will satisfy the target market segment and the business.  This is the main output from Creating Offerings.

Explanation

By this point, we should already know the market segment that we are serving.  Market segment equals job + job executor + constraint on performing the job.  If this is not known, then go back and perform this step.  

You should also know the detailed requirements for the offering.  The offering requirements are not yet set in concrete due to the iterative and learning nature of the innovation process.  We may be in the first iterations, and do not yet understand the features that will fully remove the market and business constraints.  We may still be learning where we can attack the constraint that the market segment is saddled with.  Should we do this in the product, the service, the manufacturing process, the business model or at some point in the value chain?  We are also still learning the features of the product or service that allow us to simplify the job for the market segment.

Example—Drying Shoes

There is a constant need to wash and dry athletic shoes but this continues to be a difficult job to perform.  Many people will try to get this job done by throwing the shoes into a clothes dryer. 

The shoes may not dry entirely during a typical dryer cycle.  The shoes will bang around in the dryer causing noise and may damage the dryer drum, the other clothes or the shoes themselves.  We are not constrained to use a clothes dryer, but we need to get the job of drying the shoes done. We would like to have a compelling offering for people who want to perform the job of drying shoes.  We would also like to have a compelling offering for a potential licensee.

Recall the offering requirements that will satisfy the target market segment and the business.  This is the main output from Creating Offerings.

The features of the offering are the ability to dry the shoes uniformly in 20 minutes with no damage to equipment or shoes.  The offering should cost no more than $5 dollars and should be simple to buy, install and maintain.

Method

Step 0:  You should already know a lot about commercially available alternative competitive products and services.  If you have not done a search of alternative competitive systems, then go back to the book Creating Offerings and do this now.  Because there are no competitive products or services that look like what you are considering does not mean that you do not need to look at patents.

Step 1: From patents and literature, study the history of the functions that are typically involved in the job. What functions have been added over time? What main physical parameters have improved?

Step 2: From patents and literature, study the history of the technologies (physical phenomena) that typically deliver these functions.  How have these technologies changed?  What physical phenomena and scientific effects have been used to deliver the functions?

Step 3:  Summarize the history.

Explanation

Hopefully, you have already looked at competing systems in the book Creating Offerings.  If you have not done this then you should probably go back and do this.

Now is the perfect time to perform a patent search.  Why would we do this now?  As discouraging as this may be, it is important to see how others have delivered the functions that your system will deliver.  Don't worry that your great idea has already been patented.  There are many ways around patents and patents can often be improved upon.  Also, what may be even more important is inventions that allow for the effecient manufacture of the product.  If the product does not yet exist in the market place, there may be a bad flaw that needs to be overcome.  Finally, it is best to not create a product that is a duplication of something that is patented unless the patent is no longer being inforced.  

Knowing the history [9] of a system helps in understanding the main evolutionary trends.  Each system has a main evolutionary tendency.  The tendency of a system to stall along this evolutionary path is largely a function of the technical problems that directly conflict with this evolutionary tendency.  You have already conducted a patent search within your industry so you have a lot of information about the history.  This step can take time, but the information is extremely valuable from the viewpoint of continued steps.  The inventor is becoming a true expert in this field.

Example—Measurement of Corrosion

Step 1: From patents and literature, study the history of the technologies (physical phenomena) used for detection.  How have these technologies changed?  What physical phenomena and scientific effects have been used to deliver the functions?

• From http://www.corrosion-club.com/historyelectrochem.htm Electrochemical Corrosion Monitoring - Historical Items:  1942 Hickling - Potentiostat:  The term "potentiostat" introduced by A. Hickling, who was also the designer of such an historical electronic model. An article on the web site www.bio-logic.fr also credits Hickling with the invention of the three electrode potentiostat (facilitating automatic control of the cell potential) - "genius idea" and principle still in use today. Potentiostats became widely used as laboratory corrosion measurement instruments in the 1960's.  Source: H.S. Isaacs: "Aspects of Corrosion from the ECS Publications", Journal of the Electrochemical Society", Vol. 149, No.12, pp.S85-S87, 2002. (An ECS Centennial Series Article).  1957 M. Stern and A.L. Geary:  Measurement of general corrosion rate by inverse of polarization resistance, from potential and current measurements near the free corrosion potential. Widely used in d.c. electrochemical corrosion monitoring instruments.  1960's Epelboin:  Development of electrochemical impedance spectroscopy (EIS), a.c. corrosion measurement tools.  1968: Warren P. Iverson—Measurement of voltage transients (fluctuations) in corroding metals and alloys.  "... he (Warren Iverson) published a short article in the Journal of the Electrochemical Society that many still quote as the first paper in electrochemical noise ..."  Source: F. Huet, A. Bautista and U. Bertocci: "Listening to Corrosion", The Electrochemical Society Interface, Winter 2001, pp.40-43.
  
• 1956—US2947679—corrosion rate is measured by a correlation between the corrosion current and the current necessary to lower the relative potential of the corroded metal a certain amount. It is interesting that instead of measuring the amount of corrosion, this method measures the corrosion rate which is an instantaneous measurement that could speed up the process of corrosion measurement for our example.  The method is straight-forward and allows for a nearly direct voltage measurement.
• 1956—US2856495—In order to determine the corrosion of pipes without disassembly, the resistance of a wire inside of the pipe is measured. The metal wire represents the metal of the pipe and is long and thin to allow for the maximum change in resistance due to the corrosion.  This could also be used in our example if the capsules were no longer capsules but were very long and very thin wires that would rapidly register a resistance change to the acid. 
• 1959—US3069332—Direct measurement of the corrosion current. This is related to US2947679 above.  This provides a faster was to capture a voltage representing the corrosion current.
• 1964—US3337440—Similar to US2947679 except that it discusses measuring the corrosion current when surrounded by jet fuel, etc.
• 1965—US3684679—Similar to US2947679
• 1966-US3449232--Similar to US2947679 except it provides a means of measuring the current in more difficult situations where the corrosion is caused by a drop of water in an oxygen environment.  The cathode is then formed at the interface between the drop, the oxygen and the corroded metal.
• 1966—US3539808—Irradiate sample with beta nuclear radiation and then measure back scatter and correlate with thickness of corrosion. This measurement will require several transformations to get voltage.
• 1970—US3609549—Similar to US2856495 in that the corrosion is indirectly measured by noting changes in resistance. This invention seeks to solve the problem of heating of the specimen by the resistance measurement system.
• 1971—US3748247—Similar to the 1956 patent US2947679. This invention seeks to make a convenient probe.
• 1972—US3821641—Introduces oscillating current probes and a balancing circuit to keep the current amplitudes constant. The means to keep the current amplitude constant is then measured to derive the rate of corrosion.
• 1974—US3947329—Similar to the 1956 patent US2947679 but extends the method to high rates of corrosion by imposing electrical conditions for short periods and then watching how the system reacts.
• 1977—US4130464—This is a fundamentally new method which measures the rate of decay of the potential of the electric double layer. A charge is driven into the electric double layer and the potential is reduced by the corrosion.
• 1979—US4238298—Apply alternating currents between test plates and measure the alternating current impedance. The corrosion rate is a function of the alternating current impedance.
• 1983—US$575678—The corrosion rate is measured by measuring the low frequency noise voltage between two electrodes immersed in the corrosive environment. The voltage is averaged to give the corrosion rate.
• 1987—US4800165—is essentially the same as US4238298 applied to large structures in a corrosive environment.
• 1987—US4758312—an electrochemically active material is placed into the liquid corroding environment having a reduction potential more positive than the corroded material. The material is then removed from the environment following the reaction and analyzed.   The resulting analysis is correlated to the actual corrosion of the material.
• 1990—US4006786—is essentially the same as US4238298 except that impedance is measured at several fixed frequencies and the correlated impedance is then correlated to the corrosion rate.
• 1993—US5519330—introduces a pulsing instead of different frequencies. The pulses are of different amplitudes.  A corresponding current is measured and correlated to the corrosion rate.
• 1998—US6132594—This is similar to many other methods that measure the voltages involved in corrosion except that it acknowledges that many localized corrosion can occur. For this reason, the corroded surface is segmented and many measurements are taken.
• 1999—US6275050 detects by injecting a signal into a junction and then detects 3^rd Harmonics in particular. This is interesting, but not related to our problem as we are not concerned with measurement of junctions.
• 1999—US5905376 detects corrosion using MRI
• 2000—US5320395—is essentially the same as US4238298
• 2000—US6556001--is essentially the same as US4238298 except that it injects a noisy signal that changes over time. The resulting current is examined over time and mathematically manipulated to give the expected corrosion properties of the material.
• 2002—USH2127—detects by pulse heating a surface and then watching the phase of the response. This allows to see corrosion under a coating.
• 2003—US7057177—Finally, a new method. Here, infrared radiation is reflected off of a surface and the amount absorbed is determined.  These two values are correlated to the amount of corrosion.
• 2003—US20030146749A1—Another new method that uses a magnetometer to measure the magnetic fields of the corroded magnetic materials.
• 2005—US20050274628A1—Essentially the same as US6132594
• 2005—US20050213430A1—A new method. Scan the surface of an article to the level of micrometers before corrosion and then after corrosion.  Measure the difference at each point and statistically evaluate for the corrosion rate.
• 2006—US7388386—Uses a multitude of small plates of varying thickness. Each corrode through and break an electrical path.
• 2010—US20110067497A1—Uses ultrasound to measure the thickness of the sample to determine the amount of corrosion.

Step 3:  Summarize the history.

Note that not all cases require the investigator to perform a patent search.  We are conducting one in this case because we are interested in creating new and novel ways of measuring corrosion.  Here is a summary of what was found:

• Pre 1956: Weight loss coupons—This is the oldest method.  It consists of accurately weighing a sample (coupon) and then placing it into the corrosive environment for a specified length of time.  After the time, the coupon is removed from the corrosive environment and cleaned of the corrosion.  The coupon is then re-weighed and the rate of corrosion is determined from the difference in weights and the time in the corrosive environment.  Weight loss coupons are the type of system that is being used in the acid container problem.
• 1956: Electrical resistance—Some of the first patents involved placing wires of the materials to be tested into the corrosive environments and then measuring the resistance of these wires over time as the cross section decreased due to corrosion.    
• 1956: Linear polarization—This method involves reversing the galvanic voltage and measuring the resulting currents.  This was developed around the same time as measuring electrical resistance and measuring the galvanic current.
• 1959: Galvanic current—involves measuring the actual currents involved in the corrosion process.  These methods came about at about the same time as the electrical resistance methods.
• 1966: Irradiate sample with beta nuclear radiation and then measure back scatter and correlate with thickness of corrosion.
• 1977: Measure the rate of decay of the potential of the electric double layer. A charge is driven into the electric double layer and the potential is reduced by the corrosion.
• 1979: Apply alternating currents between test plates and measure the alternating current impedance.  The corrosion rate is a function of the alternating current impedance.
• 1983: Measure the low frequency noise voltage between two electrodes immersed in the corrosive environment.  The voltage is averaged and correlated to give the corrosion rate.
• 1999: Detect corrosion using MRI
• 2003: Infrared radiation is reflected off of a surface and the amount absorbed is determined. These two values are correlated to the amount of corrosion.
• 2003: Use a magnetometer to measure the magnetic fields of the corroded magnetic materials.
• 2005: Scan the surface of an article to the level of micrometers before corrosion and then after corrosion.  Measure the difference at each point and statistically evaluate for the corrosion rate.
• 2006: Use a multitude of small plates of varying thickness.  Each corrode through and break an electrical path.
• 2010: Uses ultrasound to measure the thickness of the sample to determine the amount of corrosion.

Each of these methods increased in precision and in the ability to operate over a wide range of operating conditions as evidenced by the patents.  Most of these methods can be used to generate an instantaneous corrosion rate.  In some cases, the level needs to be numerically differentiated to give an instantaneous corrosion rate.

Incrementally and Iteratively Creating the Product or Service
The process of creating a product or service from scratch is one of incrementally and iteratively adding functional sub-systems to a system until the system is complete.  This is necessary because of the difficulty of thought in creating large systems.  While humans are advanced, we are not that advanced!  Almost all large systems are composed of subsystems that were designed and built by teams.  These subsystems are likewise composed of further subsystems that require a lot of time to design, prototype and shake the bugs out.  It may come as a surprise to many people that this is how large and complex systems are built.  Even less complicated systems are designed and built this way.

Method

Step 1:  Identify several solution concepts that meet the super-system targets and ranges. 

Step 2:  Iteratively develop each system to match the targets and ranges.  Communicate the capabilities upward to the designers of the super-system.  The idea is to continue to reduce the risk of each system while keeping the largest set of solutions possible.

Explanation

Set-Based System Creation
If your effort is large enough to consider multiple solution concepts, you may consider several system candidates at a time.  This is sometimes referred to as set-based system development.  The effort begins with several candidate systems which come close to meeting the original target and range requirements.  The capabilities of these systems are communicated upward to the super-system designers.  Over time, as the super-system designers see the capabilities of the various sub-systems, the targets change and the ranges of what is acceptable begin to narrow.  All this time, the candidate systems are evolving toward the targets and ranges given.  Finally, the ranges narrow to the final values and tolerances which the system will use.  What has just been described is the process that Toyota uses to develop its cars.  Rather than sending down one rigid set of requirements, targets and ranges are flowed down and capabilities flow back up.

Note that this all allows for prototype development, but at a larger scale.  This may not be possible for small organizations or individuals.

Also note that the design and prototype process is iterative and based upon discoveries that occur as the system is being created or evolved.  The idea is to reduce the risk of each system while keeping the largest set of potential solutions for a long as possible. 

Each solution, in a set based approach, is evolved in its own right.  Some solutions may be revolutionary with completely new physical effects and phenomena.  Others will already exist and evolve toward the given targets and ranges.  It is understood that individuals or small teams may not be able to develop multiple solutions.   

While paring down the solutions, it is important to avoid certain types of solutions.  For example, solutions that cannot be discriminated by the target market are not going to be winners.

Also avoid solutions that will lead to highly disruptive situations unless you are prepared to do battle with your organization. This may seem somewhat narrow-minded, but experience shows that most inventors cannot fight a product battle on one front and a cultural battle on another front.

When it comes to selecting the system that will be used, try to select for the simplest system with the fewest components.

Method

Step 1:  Procure a hardbound journal and ball-point pen.  Number the pages. Begin each entry on a new date with the date.  Be very careful not to go back to previous days to annotate.  Always write in ink.

Step 2:  In your Journal, identify the main useful function.  This is composed of the product that we are modifying and the modification that we are performing.

Step 3:  If all you have is the main useful function then start with this.  If the main useful function is an informing function then start with this.

Step 4:  If the function that you are adding is a useful function then go to TRIZ Power Tools—Idealizing Useful Functions and use the approaches there to Idealize the function that you are adding.

Step 5:  If the added function is an informing function then go to TRIZ Power tools—Idealizing Informing Functions and follow the algorithm at the required level.

Step 6:  As problems arise, use the methods of TRIZ Power Tools—Resolving Problems to resolve these problems.

Step 7:  Perform internet, YouTube, patent and store searches for existing systems that perform all or part of what you want your system to perform.  Look for system elements that can meet your cost requirements.

Step 8:  On each iteration, determine whether simplification is required yet. If so, then go to TRIZ Power Tools—Reducing Offering Burdens

Explanation

Where to Start:  The Primary Useful Function
This iterative process begins with the primary useful function.  The primary useful function acts on the element of the super-system (job) that it is meant to serve. This product holds a special place as the system product.  Everything that operates on this product has higher functional rank in the system.  Supporting elements which do not serve this product are generally referred to as auxiliary functions.  These have lower rank.  As each subsystem is added, the supporting elements are added further and further from the system product.  Supporting elements should be cautiously added as they are less necessary and drive the cost and complexity of the final offering.

In Summary, the process of creating a product or service from scratch is one of iteratively adding functional sub-systems to a system.  This iterative process begins with the primary useful function of the system and knowledge of how well this function has to be achieved.

Determining the Ideal Sub-system elements
One way to cautiously add subsystems is to idealize each one before we add it.  As we add each subsystem, we must manage important decisions concerning the purpose of the elements, the physical phenomenon or effect that will deliver the function and the functional objects that will deliver the physical phenomenon.  These decisions are very important and affect all downstream decisions.  Consequently, it is important that we add these elements in the most ideal way possible.  Detailed process steps are given to ensure the proper consideration.

As we add each subsystem, we must manage important decisions that will affect added supporting elements downstream.  These decisions become increasingly difficult to reverse as time goes on.  Consequently, it is important that we add these elements in the most ideal way possible. The books TRIZ Power Tools—Idealizing Useful Functions and Idealizing Informing Functions helps us with algorithms for adding functions in a more ideal way. 

In order to make the necessary design decisions, we do not take this function for granted.  We probe to see if there is something more ideal.  Is this the real product that we want to modify?  Is this the real modification that we want to perform on the product?

Idealizing Each Added Useful Function
The first step to idealizing a useful function is to identify and isolate the final ideal state in functional terms. We start by considering useful functions first, because informing functions are actually a special case of useful functions and one major path of idealizing harmful functions is to turn them into useful functions.  Once they are turned into useful functions, they may be idealized using the steps shown in this chapter.

As we add each subsystem, we must manage important decisions that will affect added supporting elements downstream.  These decisions become increasingly difficult to reverse as time goes on.  Each new function that is added requires the same process as the primary useful function.  Since this is the first and most important function that is required in the system, let’s see how the process works for the primary useful function.

The first decision that we must make is whether we really want to serve the given system product or consider a more ideal product.  What could be more ideal?  For instance, if we need to serve multiple elements, we ask if we can act upon the natural grouping or a multiplicity of the products at one time.  If the product could be slightly modified, then there might be no need to modify it at all.  If the product is a waste element then performing an upstream function in a way that does not produce the waste element may be more ideal.

Once we have determined what product we really want to serve we ask whether the modification that we have initially chosen is the one we want to continue with.  Perhaps it may be more ideal to perform the inverse modification where modifications that are relative to something are not performed, but the relative object is modified instead.  Perhaps a more ideal modification would be stated differently.  It might be in a different sequence than normally performed or done in such a way as to reduce the energy consumption.

Once we have established product and modification, then we need to identify potential physical phenomenon that will be used to deliver the function.  We usually have several to choose from.

Before we make a final choice of physical phenomenon to deliver the modification, we identify the resources at our disposal.  What systems are readily available? What have we used in the past? What elements in the system could be pressed into service? What inexpensive substances exist in the system or in the environment?  Now we are ready to identify various physical phenomenon that will be used to deliver the function.  We usually find that the potential phenomena are not unique and we may have several to choose from.

Next, we look around to decide what objects we have readily available to deliver the chosen physical phenomena.   The decision of which of the potential physical phenomena to use is highly influenced by the available resources in the system.  This is because we would like to introduce no new physical objects or substances into the system if possible. 

Following the choice of the objects to deliver the physical phenomenon, we need to decide what the architecture of the system is.  This means that we identify the attributes of the objects that have the most influence on how well the function is performed and we determine the level that these attributes must have.  These attributes and their attending levels establish the architecture of the system.  Depending on the complexity of the system, there are a variety of tools that can help us to decide the levels that the object attributes need to assume.    Sometimes, we can use models if the physics is well understood.  Other times, we may need to create simple prototypes to perform experiments.

Whether we use an existing system or not depends a lot on what our goals are and whether we are willing to evolve existing systems or we are looking for a more ideal system. Sometimes, we just want to see one part of a system working and it is ok to use ready-made system elements.  This is particularly true when existing systems are low-cost. If we are more interested in identifying much more ideal systems, then the next section will help us to do that.  There we will be looking for the most ideal sub-system elements.  Doing this preliminary search won’t hurt anything though, even if you are looking for the most ideal system.

If you are struggling with ways to deliver these functions and you have not already done a patent search from the previous steps, this might be a good time to do it.

System Simplification

Idealizing Each Added Useful Function
Before we build aesthetic prototypes (remember that they may not be required if we can show it graphically rather than physically), we may want to simplify the system somewhat.  This makes building the prototype less time consuming and will also help with the manufacturing process.

On the first few attempts, we may decide that it is not necessary to simplify in order to get an idea of whether the concept will work.  In this case, this step can be bypassed.

Example of Drying Shoes

Method

Step 1:  Procure a hardbound journal and ball-point pen.  Number the pages. Begin each entry on a new date with the date.  Be very careful not to go back to previous days to annotate.  Always write in ink.

Step 2:  In your Journal, identify the main useful function.  This is composed of the product that we are modifying and t

he modification that we are performing.

The main useful function is to move water.

Step 3:  If all you have is the main useful function then start with this.  If the main useful function is an informing function then start with this.

On the first iteration, we are starting with the main useful function having to do with drying the shoes.

Step 4:  If the function that you are adding is a useful function then go to TRIZ Power Tools—Idealizing Useful Functions and use the approaches there to Idealize the function that you are adding.

To illustrate this step, let us consider the drying of fabric shoes that become moist during washing or by accident if the water is clean enough.  The market, then, is people who are trying to rapidly dry their shoes following a wash or accident.

The ideal product is the water and the ideal modification is to move the water.  The shoes are not the product since we do not want the shoes to be modified in any way.  The moisture may have modified the shoes more than we would like, but that is a separate issue.  All that we want to do is to change the location of the moisture from being inside the shoes to being elsewhere.

Potential Physical Phenomena that could move water are:

---Evaporation

---Centrifugal force

---Movement by Gravity

---Movement of gasses

---Vacuum

---Surface Tension

---Desiccation.

Following are potential resources that could be used to deliver the physical phenomena:

---Evaporation:  Use of the sun.  The sun is not reliable in certain locations.

---Evaporation:  Simply setting the shoes out in a warm room.  The evaporation rate is normally quite low and would need to be boosted if drying time is important.

---Evaporation: A clothes dryer could be used.  Shoes bounce around in a dryer thus making a lot of noise.  They may damage the dryer or other clothes.

---Evaporation:  A food oven could be used.  Cultural biases may need to be overcome.

---Evaporation:  A microwave oven could be used.  People might be concerned that something strange might happen due to various shoe materials.

---Centrifugal Force:  The washer spin cycle has high centrifugal force, but the shoes are rarely well positioned to take advantage of it.

---Movement by gravity:  Gravity will move liquid water through a fabric up to a point but not enough to dry.

---Movement of gasses:  The clothes dryer moves air about, but the velocity of the air is low.

---Vacuum:  A vacuum cleaner can create a flow of air through the fabric.  This is an interesting possibility since clothes dryers are not really good at drawing air through fabric.

---Surface Tension:  The surface tension of the fabric in the shoe draws the water from areas that are more abundant to areas that are less abundant.  Other dry fabrics such as old towels can be used to dry off liquid, but their effectiveness is low.

---Desiccation:  The water must first evaporate from the shoes in order to arrive at the desiccant.  The desiccant is mostly a storage medium.

Finally, the physical phenomenon is chosen:  The clothes dryer is chosen since it is a natural extension of drying the clothes and the washer was already most likely used to wash the clothes.

Step 5:  If the added function is an informing function then go to TRIZ Power tools—Idealizing Informing Functions and follow the algorithm at the required level.

We are not working with an informing function.

Step 6:  As problems arise, use the methods of TRIZ Power Tools—Resolving Problems to resolve these problems.

Just throwing the shoes into the dryer is problematic.  They bounce around making a lot of noise and can damage the dryer or the clothes, if there are any.  From TRIZ Power Tools—Resolving Problems, we create a function diagram of the problem.

Note that everything in our system is red.  The green and yellow are super-system elements.  The yellow is the system product, the object that our system modifies. 

The next step is to do a causal analysis of the problem and discover the deeper problem.  In this case we will use a simplified approach and identify the harmful functions as the problem.  We note that there are two problematic functions dealing with the shoes.  The shoes damage the drum and the clothes.

Next we jump to idealizing these harmful functions in TRIZ Power Tools—Idealizing Harmful Functions.  Ideally, the shoe should not exist.  Since the shoe must obviously exist, there is a contradiction that we see can readily be resolved by separating in space.  The shoes must exist in a location where they cannot damage the clothes and drum and not exist anywhere else.  They must exist in a location where they are not bounced around but still get a good flow of air.  There are two locations.  One location is where the air enters the drum through a grouping of small holes.  Another location is where the air exits the dryer.  The air at the entrance is somewhat dryer than the air at the exit, but it may be easier to direct the air at the shoes at the exit.   Note that the harmful functions in the function diagram are gone.

Step 7:  Perform internet, patent and store searches for existing systems that perform all or part of what you want your system to perform.  Look for system elements that can meet your cost requirements.

Internet searches for drying shoes shows different approaches to drying shoes.

1—hang the shoes in the dryer by the laces.  This can be done by a variety of methods such as shutting the laces in the dryer door or special suction attachments inside the dryer.

2—Place the shoes on a dryer rack made specifically for the dryer. (This is a higher cost solution if the dryer does not come with such a drying rack).

3—Place the shoes inside of a pillow case stuffed with other cloth articles.

4—Use a hair dryer

5—Hang them on a clothes line (slow)

6—Cram a towel into the shoes

7—Stuff shoes with newspaper (slow)

Step 8:  On each iteration, determine whether simplification is required yet. If so, then go to TRIZ Power Tools—Reducing Offering Burdens.

Simplification not required on this iteration.

Method

Step 1:  In your Journal, draw and re-draw an architecural guess/design in your invention journal multiple times. Be as specific as possible. (The system may already exist but in flawed form.)  If you are not changing the system, then the existing system will serve as the baseline system. 

Step 2:  Consider witnessing the drawings.

Step 3:  As problems arise, use the methods of TRIZ Power Tools—Resolving Problems to resolve these problems.

Explanation

Drawing the Architecture (Schematics, etc.)
Once we think that we know the ideal sub-system elements, we are ready to draw pictures that show the architecture of these elements.  These drawings can be quite crude and do not need to always show scale.  They show, schematically, how everything works together.  The first drawings may be done on a napkin, so to speak or they may be drawn in your invention log or journal.  The main thing that you are trying to accomplish is to show the features of the sub-system elements that will control how well the subsystem works.  Also shown is the spatial relationship of the elements to each other.

When you start to draw the system you will begin to note various problems.  It is usually necessary to start solving these problems at this stage.  The book TRIZ Power Tools—Resolving Problems covers this subject.

This drawing process is often iterative.  The intent is still to make the added sub-system as ideal as possible in terms of the use of readily available materials.  One way to accomplish this is repeatedly drawing the subsystem, each time making small improvements. Drawing the subsystem many times can be helpful in reducing the number of elements and their interfaces.  For the final drawings, some people like to draw it as “richly” as possible.  This means adding as much visual detail (such as shading, etc.) as you can to help make the idea real.

Almost immediately, problems will arise and need to be resolved.  The book TRIZ Power Tools—Resolving Problems can be used to help to resolve these problems.

Example—Drying Shoes

Step 1:  In your Journal, draw and re-draw an architecural guess/design in your invention journal multiple times. Be as specific as possible. (The system may already exist but in flawed form.)  If you are not changing the system, then the existing system will serve as the baseline system.

On the first iteration, the drawings are simple but should probably be noted in the invention journal.  This establishes the idea and the date that it was created.  In this case, the drawings are witnessed because the idea may be significant.

Step 2:  Consider witnessing the drawings.

No need in this case.

Step 3:  As problems arise, use the methods of TRIZ Power Tools—Resolving Problems to resolve these problems.

At this point, we can anticipate problems, but it is better to prototype and see if these problems really exist.

Method

Step 1:  Create a list of design risks.

Step 2:  Identify the biggest risk.

Step 3:  What is the simplest prototype or model that we can do to to verify whether there is a problem?

Explanation

Learning Versus Execution
No matter how much we think we know, nature has a way of of teaching us that we don't know everything.  Over time, you will get a sense of where the risky issues are.  Tackle those first.

Method

If possible, model the risky subsystem.  Use the model to set parameter values.  Consider using TRIZ Power Tools—Mobilizing Function Resources to identify clever means to mobilize these resources.

Explanation

Modeling the Subsystem
Now that we have an architectural guess, we can model the system.  The need for modeling comes from trying to get the physics right or close on the first rounds.  Often, the size of the effects and the sizing of the components will determine the next level of drawing or the layout.  The outcome of the modeling can also effect the basic schematic that we have just made if the sizes of the elements are much different than we originally expected.

Models are usually created from a functional drawing of what is going on such as a timing diagram or a free-body diagram.  Beginning models should be as simple as possible to give a rough-order-of-magnitude of the sizing.  The need for more detail will become evident on further iterations of the prototype. 

It is realized that many people do not have the background to model the physics.  If this is the case, then this step may be skipped in favor of jumping to the prototype, but the risks of spending a lot of time on prototyping grows substantially.  An example of an inventor that avoided modeling was Thomas Edison.  Nichole Tesla worked with Edison for a time and noted that Edison’s operation would be greatly improved by employing a little physics.  Tesla was a noted physicist in his day and had little tolerance for cutting and trying.  Edison, on the other hand, was well noted for his large-scale invention factory where people were employed to cut and try.

Why Model?
The need for modeling comes from trying to get the physics right or close on the first rounds.  Often, the size of the effects and the sizing of the components will determine the next level of drawing or the layout.  The outcome of the modeling can also effect the basic schematic that we have just made if the sizes of the elements are much different than we originally expected.

How are Models Created?
Models are usually created from a functional drawing of what is going on such as a timing diagram or a free-body diagram.  Beginning models should be as simple as possible to give a rough-order-of-magnitude of the sizing.  The need for more detail will become evident on further iterations of the prototype.

What If I Don't Have the Background to Model?
It is realized that many people do not have the background to model the physics.  If this is the case, then this step may be skipped in favor of jumping to the prototype, but the risks of spending a lot of time on prototyping grows substantially.  An example of an inventor that avoided modeling was Thomas Edison.  Nichole Tesla worked with Edison for a time and noted that Edison’s operation would be greatly improved by employing a little physics.  Tesla was a noted physicist in his day and had little tolerance for cutting and trying.  Edison, on the other hand, was well noted for his large-scale invention factory where people were employed to cut and try. Sometimes, cutting and trying is just too slow.  Even simple models of the sub-system can save a lot of time, especially if there is a lot of physics involved.  Models can help you to set the levels of the controlling features which, in turn, helps to visualize the final system.  If the physics are not well known then it may be good to hypothesize the physics and then build a model using the hypothesis as a guide.  As a matter of practicality, the author has often found that many conceptual models can be built by going back to “first principles” such as conservation principles, continuity, etc.

Rapid Prototyping Versus Modeling
Sometimes, rapid prototyping is much faster than modeling so this step can be skipped.  This is often true in cases where setting up and running models can be very time consuming such as computational fluid dynamics models.  On the other hand, be open to advances in areas that have been difficult to model.

Example—Cleaning Shoes

If possible, model the risky subsystem.  Use the model to set parameter values.  Consider using TRIZ Power Tools—Mobilizing Function Resources to identify clever means to mobilize these resources.

On the first round, we are jumping directly to the prototype and don’t need to do this.

Method

Step 1:  If possible, create scale models or drawings of the subsystem that we can build/prototype.  It may be necessary to do this several times in order to create a simple architecture.

Step 2:  Build simple prototypes and perform experiments to understand the magnitude of the effect that you are interested in and what structures work to deliver the intended functions.

Step 3:  If you discover that the physics is different than predicted, observe what features really control what is happening.  Is the physical phenomenon what you thought it would be, or are you dealing with something different?

Step 4:  If you have the ability to try different configurations rapidly, then do this to get as much understanding as possible of the features that control the performance.

Step 5:  Iteratively resolve problems that you may run into.  If the problems are serious enough to stop you then refer to TRIZ Power Tools—Resolving Problems.

Explanation

Whether you model or not, it is important to perform actual experiments and/or build prototypes to verify that your understanding of the physics is correct.  So often, I have found that I missed something.

Build to Think and Think to Build
Building and testing functional prototypes helps us to see the physics and think deeply about what is happening.  IDEO[1] refers to this as “Build to Think and Think to Build”.  This is where the greatest learning occurs because we get to see how various parameters of the objects control the functions that we are trying to deliver. Humans have the innate ability to think with their hands[2] . There is no substitute for making stuff yourself.  The physical act of creation teaches us in so many ways.  This is where intuition comes from, practical experience.  The more senses that get involved, the more that we will know when we are finished.  This type of knowledge cannot be learned from textbooks and is often not valued in the educational world.  When we act and experience, our brains record thousands of pieces of “tacit information”.  The skilled craftsman is filled with such knowledge and so are the people that build their own prototypes, no matter how crude.  Salvaging pieces of equipment from other devices and using them again only increase our tacit knowledge.  While these books are trying to help us to understand other intellectual tools for inventing, we also need to increase the skills and experience necessary for intuitive inventing. Even though these books are to help us move forward intellectually, there is an element of inventing that cannot be replaced with technical knowledge and that comes from building and taking apart stuff.  Working with others that have this tacit, practical knowledge can save a lot of time, especially when it comes to discovering how to manufacture your ideas.

Prototyping also increases our personal data banks for future projects.  Our ability to prototype grows with each project as does our tools and raw materials.

How Dumb we Really Are
Sometimes humans think that they are smarter than they really are.  It is not unusual for an innovation consultant to help someone come up with a "Break-through" idea only to find that there are a lot of problems left to be solved.  Prototyping is where we may discover difficult problems such as harmful functions and contradictions.  These problems should be handled at this step.  TRIZ Power Tools—Resolving Problems helps us to overcome these problems in a systematic way.  This is an iterative process that can take some effort and time for each subsystem.  This step can have interesting effects on the emotions and intellect of the innovator as we see how our hypothesis works out.  Sometimes discouragement sets in and we need to take a short break to think through what we are doing.  This is fine so long as we do not let the discouragement stop us.

The Value of Simple Prototypes
Many product developers get caught up in prototyping and spend too much time and resources here.  IDEO will be the first to tell you that starting crude and simple is the best way to go.  One can learn a lot from simple prototypes.  The purpose and audience of prototypes is discussed further.  Remember that the prototype should be matched to the audience and the product developer is the most important audience to begin with.  You need to be convinced that your idea is worth its salt before you have the courage to continue.  Following is a great example of the spirit of prototyping by Patrick Hale and was given at the Mark Niver memorial.  "My name is Patrick Hale . . . and I want to share a story with you about a very unique friend of mine, who was known for giving and sharing in the most unusual of ways. Let me take you back in time . . . a time when Jimmy Carter was in the White House.  I started drag racing about the same time Mark Niver did. I made my very first pass at Beeline drag way in 1970 during my freshman year at ASU. A few years later, in 1978, I bought my very first dragster from Mike Abby, who along with John Powers, are our hosts here tonight. Thank you guys for this tribute to Mark Niver. 

At the time I was a young engineer, fresh out of college, working at AiResearch. I had also started working on a computer program that could predict the ET and MPH for all kinds of drag racing cars. This computer simulation worked really well for my own dragster and a few other racers knew about it at the time.

At the end of 1979, another AiResearch engineer tracked me down and asked if I could do a computer simulation for a Pro Comp dragster. He had an idea for some new kind of supercharger and wanted to know how much quicker and faster a dragster would be if the standard 14-71 blower was replaced with a “Lysholm compressor”. That engineer was Norm Drazy. I told Norm . . . yes, I could do that. But I would need to know some things about a Pro Comp dragster, like how much it weighed and how much HP it made. As we got more and more into the computer simulation the aero differences between my small block Chevy dragster and a big-tired, big-winged, Hemi-powered Pro Comp dragster needed to be understood. This was early 1980, there wasn’t any wind tunnel data for dragsters, and my experience with Lear jets just didn’t apply.

Drazy said he would arrange to get us some data. I was very excited about all this – as engineers just love data. A couple of weeks pass, and Norm calls me at lunch . . . it’s Wednesday, March 12th 1980, over 30 years ago. Norm says “we’re finally ready”. He wants me to ride my 750 Honda down the freeway to Riggs Road south of Phoenix, on the way to Casa Grande. I asked him “how far south is Riggs Rd”, as I’ve never heard of it. Drazy says it’s about 8 miles south of Maricopa Rd. This is long before Charlie Allen ever dreamed about leaving OCIR and building Firebird.

So after work I get down there about 5:30, take the Riggs Rd exit off I-10 and what do I see . . . a full-on Pro Comp dragster sitting on the side of the road. Norm Drazy is hooking up a long rope from the dragster to back of his little blue Datsun pickup truck. Who’s dragster is it? Mark Niver.

Well, what are we about to do? We’re going to tow Mark’s dragster down Riggs Rd . . . a two-lane public road on the Indian Reservation, in the middle of the afternoon with traffic going back and forth in both directions and measure how much power it takes. I think to myself . . . what a brilliant idea! The Indian police should be here any second.  Quickly, Mark yells “let’s go” and jumps into the dragster, ready to go 100+ MPH, all excited and eager to help with this “Crazy” Drazy science project. Norm gets behind the wheel of his little truck and tells me to jump in the back to read the beam scale he has rigged up to measure the force required to tow Mark’s dragster.  We start out at 20 MPH and it takes about 45 pounds of force. I write this down in my notebook using the mechanical pencil from my pocket protector – after all I had just come from work. Then we go 30 MPH, 40 MPH . . . 50 MPH. Mark’s giving us the thumbs up, yelling faster, faster . . . let’s go . . . as he’s used to being in the seat at over 200 MPH. But the wind’s really blowing me around in the back of the Datsun at only 60 MPH and I’m about ready to call this whole thing off.  But we’re getting good data . . . real data from a real Pro Comp dragster. I’m just as excited as Mark and Norm, we’re all young and the adrenaline’s flowing. By this time we’ve traveled several miles west down Riggs Rd, so Norm decides it’s getting late and floors the 4 cylinder pickup and we slowly start picking up more speed. This long . . . choo-choo-train . . . of a Datsun pickup truck, extra-long tow rope and Pro Comp dragster hits 65 MPH and that’s it. Steam’s pouring out from under the hood of the Norm’s little pickup.  Norm and Mark bring everything to a safe stop. Mark can’t stop talking about wanting to go faster – 100 MPH he shouts! We cool off a bit . . . turn around and head east, back toward Mark’s waiting trailer. On the tow back we try again to get past 65 MPH but the little Datsun just doesn’t have the HP to do it.  Finally, we load the dragster back into Mark’s trailer and take off in three different directions. Life is good . . . I have great data . . . nothing’s broken . . . no traffic accidents . . . and best of all . . . no jail time with the Indian police.  And I wonder to myself on the ride back home to Tempe . . . what kind of a man let’s two crazy aerospace engineers tow his dragster as fast as they can down Riggs Rd on a Wednesday afternoon? Mark Niver".[3]

Prototype at the Level You Need
Each level of prototyping can be used to answer important question.  In general, one should work from extremely crude and simple prototypes to more complex ones.  In the early stages, you are the audience of the prototype.  You are the one that needs to be convinced.  Seeing the effect of a prototype can be both encouraging and discouraging.  Perhaps you expected to hit a home run and what you got is a bunt.  It can be discouraging.  Other times, you may find that the effects that you are looking for exceed your expectations.  More often than not, you discover that the physics are very different than you thought they would be.  Since you had not anticipated this, it can be discouraging.  This can be true, even when you have worked in an area for many years.  What is important is that you don’t give up too early.  Get over your “failures” as quickly as you can by learning from them.  When you learn that the physics are different then the next question should be: what are the features and feature levels that control what is happening.

I will give an example from my hobby of gardening. I am very interested in vegetable gardening in the desert.  When you garden in the Arizona desert, there is little margin for error when it comes to making sure that your plants are correctly watered.  In the winter time, the plants that grow during this season usually like to be watered from above.  People who come to the desert from other climates often forget the benefits of overhead watering,  for instance, the elimination of aphids and thrips.  My garden boxes are about 4 or 5 ft across and are above the ground about 3 feet. 

While it seems  improbable, there are very few sprinklers that work well for overhead watering in a garden of the sort that I tend.  Minimizing over-spray is important to conserve expensive water.  I have been unable to find commercial sprinkler heads that met my needs so, I have turned to making my own.  Some of the early prototypes were designed to look like a shower head.  These worked very well, but the small streams that came from them did not randomly cover the plants.  This random covering is important for removal of insects.  A way was sought to create a random watering pattern.  The methods of this chaper were roughly followed.  In trying to idealize the function, I sought a physical phenomenon for spreading the water that would allow for a random pattern.  Eventually, I came to the idea of shooting a stream of water toward a target or allowing drips to strike a target, thus splashing out in a pattern.  I hypothesized that the shape of the target would allow for different patterns of spray.  I was able to find a commercial sprinkler that could be modified to provide a stream of water and a target in the same fixture.  The stream was quite narrow but sufficiently large to avoid plugging from calcium deposits.  This was just as required, but a problem arose when the target was built and tested.  The water did not come off of the target in the way that I had naively imagined.  When the water was turned down low enough to give the required size of spray pattern, the water “hugged” the target.  I had seen this before and recognized what is referred to as the coanda effect.  Fluids tend to hold to surfaces longer than expected and can sometimes shoot off of the surface at surprising angles.

At first, I was discouraged and set it away for a couple of days.  On a whim, I later tried different sizes of targets and in doing so, discovered that smaller rounded targets did something very different.  The water enveloped the targets and flowed from the opposite side.  As the water came together on the opposite side, it formed drops that came off and sprayed into the air in more predictable ways.  The drops were random, but the overall pattern was stable and the coverage was the right size.  This effect could now be used to create unusual patterns to cover unusual shapes of garden boxes.  The controlling features were now becoming more apparent.  The end of the story has not been written as prototyping continues.

Note from this example that there is a certain uncontrolled feel to it.  That is why we do prototypes.  We do it to learn and we don’t know exactly what is going to happen.  While it can be discouraging, it can also be exhilerating when one discovers things that others may have missed.  Have fun!

Create the Crudest Prototype Possible to Get the Job Done
The functional prototype simply gets the job done.  In some cases, it just demonstrates the basic phenomenon of critical aspects of the system.  Often, it is not necessary to develop a polished prototype in order to sell the idea, especially if you are the audience.

Many people put too much cost and effort into building prototypes.  Often, a crude and expensive prototype is more valuable than an expensive and finely tuned one.  The act of building the prototype causes us to think and see things that we would normally not see.  Crude prototypes often show us what we need to know.  Thinking about how to build the cheapest prototype possible is an inventive adventure and often requires mental effort. Remember that in the first round of prototypes, the audience is usually you.   This prototype will give you the knowledge and ultimately the courage to carry on.  Be willing to make several rounds, improving each round.

Crude Dinensioned Drawings
Now that you know the levels of the parameters, it is possible to make scale models or drawings of the subsystem.  These drawings will greatly help you to see how things fit together and potential problems that can arise.  They will also give ideas as to how the sub-system objects can more conveniently fit together.  

This is another step that can be helpful, or it can get in the way of crude prototypes.  (On the first rounds you may want to skip this step in order to create very crude prototypes.  Sometimes just slapping together a bunch of stuff is all you really need to do).  However, in order to preserve materials, it is often helpful to draw, at least, a hand drawing using dimensions in order to purchase the right amount of materials and to make sure that everything fits together properly.  The skill to know when to draw detailed drawings will come with experience. Most of the time, it is not necessary to make a highly detailed drawing which follows drawing standards.  If you are working with a modeling shop, it is usually sufficient to create a very simply dimensioned drawing.  You should probably only be using a modeling shop if you are on later rounds of prototyping.  In the early stages, it is often overkill.  Building your own prototype in a garage is usually faster and gives more rounds of learning than having someone build it for you.

Dealing with Harmful Functions
In the process of designing or prototyping, we may discover that we have added harmful functions.  These harmful functions need to be idealized or neutralized.  It may sound somewhat counterintuitive to consider idealizing something that is actually harmful.  It would seem to instantly create an oxymoron.  For instance, we might find ourselves considering the “ideal pain”, “ideal wear” or “ideal product failure”.  While this might sound ridiculous, we shall see that there are ways to think about this that can turn harm on its head.  In the end, harm must not exist and might even become useful.

How we handle the objects which cause or receive harm have a lot to do with what functions they perform in the system.  In general, objects which support the primary function can be eliminated, allowing their useful function to be performed by something else in the system or super-system.  If the object performs a primary function, it is difficult to remove this element and therefore we need to consider other approaches.  If the element is not required in the system at all, then it should be removed.  If it cannot be removed, then it needs to be weakened, channeled or redirected.

Example—Cleaning Shoes

Step 1:  If possible, create scale models or drawings of the subsystem.  It may be necessary to do this several times in order to create a simple architecture.

On the first round, we are jumping directly to a simple prototype and don’t need to create scale drawings.

Step 2:  Build simple prototypes and perform experiments to understand the magnitude of the effect that you are interested in and what structures work to deliver the intended functions.

The shoes are placed below the dryer vent with the expectation that it will not work very well.  The author checked half way through the drying cycle to find that they were bone dry.  It may have worked too well!  Clearly, anyone with an outside vent to their dryer can do this easily.  If I were trying to create a product, this would not be a good candidate as the target market needs nothing more than an external vent to their dryer.  This forces us to consider assumptions on previous steps.  The assumption was that the target market might need some help to overcome the constraints.  Now, the target market would need to narrow to those who are still constrained.  For instance, they do not have an external vent, or there might be shoes that still need help drying due to poor circulation of the air.

Step 3:  If you discover that the physics is different than predicted, observe what features really control what is happening.  Is the physical phenomenon what you thought it would be, or are you dealing with something different?

Step 4:  If you have the ability to try different configurations rapidly, then do this to get as much understanding as possible of the features that control the performance.

Step 5:  Iteratively resolve problems that you may run into.  If the problems are serious enough to stop you then refer to TRIZ Power Tools—Resolving Problems.

The first round is a crude and simple prototype.  The shoes are placed directly at the exit of the dryer and remain there while a load of clothes are being dried.  Below is a picture of this crude prototype. The shoes are placed below the dryer vent with the expectation that it will not work very well.  The author checked half way through the drying cycle to find that they were bone dry.  It may have worked too well!  Clearly, anyone with an outside vent to their dryer can do this easily.  If I were trying to create a product, this would not be a good candidate as the target market needs nothing more than an external vent to their dryer.  This forces us to consider assumptions on previous steps.  The assumption was that the target market might need some help to overcome the constraints.  Now, the target market would need to narrow to those who are still constrained.  For instance, they do not have an external vent, or there might be shoes that still need help drying due to poor circulation of the air.

Method

Step 1:  Have all the requirements been satisfied with the prototype?  If not, then jump back up to Identify Ideal System Elements and iterate again.

Public testing of Prototypes
Finally, we get back to the outer loop of all that we are doing.  With the prototype, we are able to go back to Book 2—Choosing Features and continue our interviewing process.  With prototype in hand, we can show our idea to potential customers and get their feedback.  We want the most honest feedback possible because we are going to sink a lot of time and money into this concept.  We may also need this type of information if we are going to license or sell the idea.

The Different Types of Prototypes
There are different types of prototypes.  The first type of prototype we have already come across if we are trying to understand a physical phenomenon.  Let’s call these the Effect Prototype.  These prototypes can be very crude and simple.

A Functional Prototype is the next step up.  Here, we are trying to see if we have the right parts and if it works in the job that we are going to use it.  Again, it is usually not necessary to develop a polished prototype in order to see the how well it works in the job.  We want to verify that there were no major risks left.  You also want to convince yourself that you haven’t left anything out so that your invention is ready for patent filing.  You may also want to use these prototypes to convince the business and some enlightened customers that your idea is marketable and can be manufactured.

Finally, an Aesthetic Prototype may be necessary in order to get a less biased reaction from the using public.  Remember that some people that may want to use your product would never consider using it if it were not “socially acceptable.”  The revulsion at seeing a functional prototype might outweigh the desire to perform the job.  These prototypes can be quite expensive.  Consequently, it may be necessary to wait for the company that is going to manufacture the product to build these prototypes. 

Crude and Simple for the First Round of Prototypes
Many people put too much cost and effort into Effect and Functional prototypes.  Making expensive prototypes can exhaust your resources. Discouragement often sets in and the final product may never materialize. Remember that the purpose of these first prototypes is to convince YOU that all of the ideas work together.  The added cost is often not justified.

There are good reasons beyond cost for making the prototypes crude.  The act of building the prototype causes us to think and see things that we would normally not see.  When we build expensive prototypes, it takes too long to see what we need to see.  The first prototypes will likely have flaws.  This is actually good because you will still learn a lot from them.  Crude prototypes often show us what we need to know both good and bad.  When we build something that does not work, we often learn of weaknesses that could crop up in an expensive proto type that is built correctly but at the extremes of its variability.

Another purpose of these prototypes is to help understand the relative importance of each architectural feature. The more rapidly you can change the features, the better because this gives you a better feel for what the knobs (attributes or characteristics) are and how much they affect the outcome.  So, just slap some stuff together and start mixing and matching.

Remember that in the first round of prototypes, the audience is usually you.   This prototype will give you the knowledge and ultimately the courage to carry on.  Be willing to make several rounds, improving a little on each round.  Even in later aesthetic versions, you are the one that needs the data and the convincing to continue.

Thinking about how to build the cheapest prototype possible is an inventive adventure and often requires a lot of mental effort.  This effort is valuable as it keeps our skills sharp and forces us to think in ways that are really helpful to the business.

Converting from Prototype to Production
While crude is often good for prototypes, a word of warning is necessary when converting to production prototypes.  The author has seen many instances where the conversion to production did not go well.  When converting a prototype to production, designers will often make seemingly small variations from the prototypes which have drastic consequences.  For instance, it is possible to hand-make prototypes in ways that are difficult to recreate in production.  Designers may not know which characteristics are critical to proper functioning of the product and make small changes believing that there is no obvious harm.  Designers may need to make concessions due to packaging limitations that disrupt critical features.  This is not an exhaustive list of what can go wrong.

In order to avoid these problems, first be aware that they can and will likely happen.  Second, build production grade prototypes as early in the development process as possible. (The foregoing discussion relates to the research and development stages as opposed to production stages).  Development testing and production testing should be conducted on parts that are as close to production grade as possible.

Method

Step 1:   Who is really the audience: the consumer, a potential licensee, an original equipment manufacturer (OEM) or your own business managers.  The first few times around, the audience is likely a potential customer that we need feedback from.  The minimum prototype should probably be physical.

Step 2:  How simple is the idea to comprehend? Remember that the more difficult the explanation, the more difficult to sell to any of the above audiences.  Licensees want the idea to be obvious to consumers.  If the idea is not obvious then you or your business may need to manufacture it.  This takes more salesmanship and consequently better prototypes.  A functional prototype might be enough to convince the business, depending upon the imagination of the critical decision makers.

Step 3:  What do you need to prove to your audience?  Is it the basic physics?  Is it the controllability? Is it the reliability or the aesthetics?  If this is the first few iterations with the customer, you are not looking at proving anything to the customer.  You are trying to find out whether the product or service plays well with the customer.

Step 4:  Determine the minimum that must be built to prove what you need to prove to your chosen audience.

Method

Step 1:  Draw multiple pictures in your journal showing several different aesthetic forms of the product.

Step 2:  Show the pictures to others and get their opinions.

Step 3:  Since you have already filed for the patent, getting professional advice is a great idea.

Explanation

The aesthetic form of the invention is often important, especially if it is a consumer product.  Notice how you react to choices that you make when you buy among competing products.  Given the choice of products having the same function, you will likely choose the one that has the most aesthetic appeal.  If your product is not competing with any product of consequence, the aesthetic form is used to convince the customer that the product will actually perform as advertised.  In other words, if the manufacturer has taken the time to refine the aesthetic form, surely the effort has been expended to ensure that it works!  Remember the user interface that you developed in in TRIZ Power Tools--Creating Offerings.

Method

Step 1:  Research the parts that must be purchased to build the prototype: motors, fans, electrical equipment, etc.

Step 2:  Create detailed drawings of the parts.  Make sure that the parts fit together.  It is usually not necessary to define detailed tolerances for professional prototype builders, but this can help the builder to determine the cost up-front.

Step 3:  Build the prototype.  Remember that you can still use many of the inexpensive materials that we have already talked about.

Step 4:  Ensure that the prototype meets all of the requirements chosen in TRIZ Power Tools--Creating Offerings.

Explanation

Once more, we build a prototype.  This is a more complete and aesthetic version.  The purpose of this prototype is to convince you whether there is sufficient public acceptance of your invention.  It is also used to determine the preferred form of the invention and perhaps the preferred manufacturing process.

Whether you create the preferred aesthetic form or not will be determined by your pocketbook.  The closer to the final product the better, as you will be able to better assess public reaction.  This is where it pays to have a real arsenal of tools if you are going to do it yourself. 

You may also wish to employ the services of a prototype shop.  This can become quite expensive and it is necessary to create detailed drawing to convey the ideas. Often this is unnecessary if you are able to develop a reasonable functional prototype and have good artistic representations of the final product.  You may want to use a 3D printer to build the prototype.

Method

Using the same steps used in creating the functional prototype iterate the manufacturing process.

Step 1:  Determine whether the methods used to prototype the product or service is indicative of the methods which will be used in full-scale production.

Step 2:  If not, then search out ways to manufacture the product for the least cost possible.  This effort may need to be as inventive as determining what the product or process should look like.  Use the same methods to solve inventive problems as you would for a product or service.

Step 3: Determine the recurring costs of the materials and labor.

Step4:  Determine the modifications to existing equipment or the requirements for new equipment.  Use this to determine the non-recurring costs.

Step 5:  Determine the non-recurring labor costs to design.

Explanation

Building the prototype really enhances the ability to understand how the product or service will be manufactured.  We say this, realizing that the prototype may not be using the final manufacturing process.  It will, however, help you to understand how accurately certain features need to be manufactured in order to make it work.

Ultimately, you will need to investigate how to produce or manufacture the product or service.  This is usually done be talking to people who are in the industry.  Don’t be afraid to investigate new ways of manufacturing.

In the previous book, we may have learned that the constraint that we want to work on, actually exists in the manufacturing process.  If this is the case, then the manufacturing methods and prototypes are the most important part of the innovation process.  We may need to focus the full innovative creativity around manufacturing the product or service.  In fact the iterative creation of the prototype may actually have been a prototype of the manufacturing process.

Understanding the manufacturing process is important if you are going to license or sell your idea.  One of the first things that a prospective licensee will want to know is how much it is going to cost and what sorts of costs she will incur to modify or buy machinery to make the product.  Consequently, it is important to anticipate every question that might be asked.

The output of this book is prototypes that we can show to the customer and to the business.  If we find that the cause of the customer constraint needs to be addressed in the manufacturing process, then the prototypes that we have already created will showcase the manufacturing process.  That is not what this step is about.  We need to understand, for the sake of the business, what non-recurring and recurring costs will be required in manufacturing the product or providing the service.  Will the business need to purchase new equipment or can the existing equipment be modified?  What will be the recurring costs of manufacturing the product or providing the service?  It is very difficult for the business to determine whether to move ahead unless this information is known.  Thus, there is a need to investigate the manufacturing or production methods as part of the prototyping process.

With the manufacturing processes understood, we are now ready to file for patents or, at least, create a place holder at the patent office with a provisional patent.  While this is not always necessary or even desirable, it does provide the maximum protection before you perform public evaluation of your prototypes.  We may want to file patents for the product and/or the production process.

Two types of patents are considered: a provisional or “poor man’s” patent and a non-provisional or regular patent.  Congress created the provisional patent in the late 90’s to help spur small business growth.  A provisional patent is basically a place holder for a filing date, since the patent examiner will not even look at it until the regular patent is filed. (It can also serve as a substitute for building working prototypes, also called “reduction to practice”).  The provisional patent must include sufficient detail that someone “skilled in the art” could build your invention.  While it is good practice to give as much detail as possible, do not worry about flowery legal wording and drawings.  Attach legible drawings and the cover-sheet provided through the USPTO web-site and pay a fee of about $80. Provisional patents are in effect until one year from the date of filing.

They need to be converted to a regular patent within the year or your invention is considered to be abandoned.  One strategy is to file for a provisional patent and then license the product or service.  Part of the business agreement is up-front money which can be used to file the non-provisional patent via a patent attorney or agent.

Traditional patents can be much more expensive.  If you are in business for yourself, it is usually difficult to justify the expense of a non-provisional patent, before the market has proven itself.  On the other hand, most large companies have greater resources and therefore rarely bother with the added time and effort of a provisional patent.

When you have established a filing date, you are now patent pending.  Write patent pending on your prototypes. You are now free to sell, license or manufacture your invention.  Understand that you are still at some risk that someone else has already patented the idea.  Your degree of risk is directly related to the depth of your patent searching and to how close your invention is to other related patents.

The subject of patents cannot be adequately covered in this text.  It is suggested that you read a book on patenting.  It is entirely possible for a lone inventor to patent his own invention and do a better job than is typically done by professional patent attorneys and agents.  However, there are a number of important things that must be learned.  Remember, you are the one that knows your invention the best.

There are a number of tests that must be overcome in order to secure a patent.  The two most important tests are novelty and non-obviousness.  The novelty test is generally an easy test.  We just need to answer the question:  Is it physically different than prior art? The non-obviousness test is much more difficult to pass.  It must be non-obvious to someone skilled in the art with full knowledge of all prior art, and it must be non-obvious after consulting all prior art. Ask yourself, does the effect of the invention provide a non-obvious result?  For example, does it solve a previously unsolved problem?  The practical rules of non-obviousness are documented to some degree in the USPTO literature.

Method

Step 1:  Document the building and testing in your journal; the more information that is written, the better.

Step 2:  Sign and date the invention in the journal.

Step 3:  Have two reliable witnesses sign and date an entry that indicates that they understand the invention and have seen the invention reduced to practice.  These witnesses should be sufficiently competent to understand the workings of the device. The witnesses should not be family members or a co-inventor.  Make sure that it is someone that you trust.

Step 4:  Indicate that the material is confidential, such as “I confidentially witnessed this invention.”

Explanation

Documenting the Idea and Witnessing the Reduction to Practice
Once we have proven to ourselves that our idea is workable, we may want to document the idea and witness the reduction to practice.  In some countries, where “first to patent” is the rule then time may be of essence and a patent should be filed.  In the United States, the first to witness the reduction to practice has the upper-hand and you may want to wait until you have added all of the sub-system elements to file.

Documenting that the invention has been “reduced to practice” is very important.  This will be required later when the patent is submitted.  Every inventor should have a patent journal and, and if possible, there should be a journal for each idea that you are working on.  Having a separate journal reduces the confusion as to which inventions entries belong to.

The entries in the journal should include as much work as possible, regardless of the prototyping stage.  In the United States, ownership is determined by first-to-practice rather than first-to-file.  Thus, the journal is the most significant proof as to when ideas were created.

In general, journals should be hard-bound with numbered pages.  You can put the numbering in with indelible ink.  You should always write and draw entries with indelible ink so as to avoid suspicion that an entry has been changed.  If you go back to change something that you wrote at an earlier date, then use a change system that you always follow.  The author will cross out and write the correct information and then date the change.  It is more secure if you rewrite the entire entry on the next available page and date it.

Any time that it is possible, you should review entries with someone that can witness the entries and sign that they understand.  Do not ask family members to sign for you.

Method

Perform a patent search to determine the patent landscape and whether your invention may be violating someone else’s invention.  If you would like more detail on this process then consult the materials related to idealizing physical phenomena in the book TRIZ Power Tools -- Idealizing Useful Functions.

Explanation

If you have not done so, it is probably a good idea to see what has already been patented to determine whether you are violating someone else’s patent.  This may have been done much earlier when trying to determine the physical phenomena that would be used.  The patent search gives a lot of information that you will need now if not earlier such as the companies that are producing similar products, etc.

Method

Step 1:  If there is a problem, then try to determine if there is a special angle that you can use to get protection.  Seek a professional if this is difficult for you.

Step 2:  If the idea is patentable then broaden the Invention

• --Broaden the Market—(Perform the same Function in more markets)
• --Broaden the Jobs it can help to do
• --Increase Functions it can do
• --Broaden the ways to implement the invention
• --Broaden Physical Phenomena
• --Broaden the Architecture
• --Value the Solution Sets
• --Prioritize the Solution Sets

Step 3:  Determine whether a patent is required for the product or service and a separate one is required for the method of producing the product or service.

Explanation

Normally, it is sufficient to file for a provisional patent before you go out to sell or license your invention.  Sometimes, the patent landscape is rockier and you may need to do more work.  If you discover that your idea is not novel in the current form, you may need to solve additional problems in order to make it novel enough.

Patent interest may go deeper than securing a single patent.  Some situations require multiple patents to create a “patent fence” to exclude competition. 

Since most examiners do not want to see a patent for a product and the manufacturing method together, you may need separate patents for these.

Method

Step 1:  Get forms from USPTO web site.

Step 2:  Fill out the forms and attach the explanation of the invention along with the drawings.

Step 3:  Include a check which is about $80

Step 4:  Include a self-addressed post card with check boxes for everything that you have sent.  They will check the boxes and return the card to you. Make sure the card has a postage stamp.

Step 5:  Send the package by U.S.  Overnight express mail.  The postage date is considered the filing date.

Step 6:  If everything is OK at USPTO then will receive filing date in 6-8 weeks.  You are now “Patent Pending”

Explanation

 If you are not sure that your idea has a strong market, or you cannot afford the expense of a regular patent application, you may want to file a provisional patent first.  Provisional patents were created in the late 90’s as a poor man’s patent.  They serve as a filing date place holder and allow the holder to be “patent pending”.  Thus, the inventor to go out and find a customer.  In other words, you can test both function and marketability in public.  Provisional patents are considerably less formal than the non-provisional patents.  They usually require a cover letter provided by the patent office and an informal explanation and drawings.  The weakness of the provisional patent is that anything that was overlooked and later put into the non-provisional patent cannot claim the filing date of the provisional patent.

Method

Step 1:  Read a good book on the subject.  One such book is Patent it Yourself by attorney David Pressman.  NOLO produces this book and it can be found at Nolo.com.  This book contains forms and in-depth explanations of the finer points of patenting.  Whether you patent yourself or allow someone else to do it, you should probably read this book.

Step 2:  Follow all of the procedures described in Patent it Yourself.  It is strongly suggested that the only procedure that should be done out of sequence is the patent search.  In this book this is done as one of the first steps.  The search helps us to determine available physical phenomena and keeps us from putting in too much work on an idea that is not likely to go anywhere.

Explanation

If you are certain that your idea has a strong market and you can afford the expense, you may want to file a regular or traditional patent first.  It is entirely possible for an inventor who is not a patent agent or an attorney to file a patent.  Nor is it required by the patent office.  A lone inventor can prosecute (do what it takes to secure) a patent from the USPTO.  On the other hand, you can also work with a patent agent or attorney.  The cost  may be  several thousand dollars.  If you are hiring someone to do the work, then follow the same guidelines as the “Traditional Patent for a Large Company”.  If you are doing it on your own then:

Method

Step 1:  Assist the company attorneys to file the patents.   This usually involves some sort of disclosure of the invention, meeting with the attorney, disclosing drawings of the idea and providing lab notebook copies.

Step 2:  Assist in the patent search.  The company will usually hire a search to be done.  The patents that come back need to be reviewed.  Some search firms are better than others.  If you have developed the ability to search patents yourself, you can assist.

Step 3:  Review the patent application for correctness.  Remember, you are the one that truly understands how your patent works.  Take the time to review it properly and give needed feedback.  Especially take the time to carefully review the claims.  Create a claim tree to help you understand the completeness of the claims.  The strength of the patent will largely depend upon the strength of the claims.

Explanation

If you work for a company, it is often the policy that a regular patent is filed first.  This is because a non-provisional patent is more complete.  It is done once and has a stronger position than the provisional patent.