1. How Inventions Come About
2. How to Encourage Innovation in Your Shop
3. Advice on Patenting
4. Advice on Start-up Businesses
5. Advice on Prototyping
6. References

The Grolier Encyclopedia (which is bundled in with Windows 95) defines invention as follows:
"...an act of creativity that results in a device, process or technique novel enough to produce a significant change in the application of technology."

An invention may be "entirely new," or it may be an application of existing knowledge(s) in a different way.

There are three types of inventions: need-driven, device-driven and accidental. The first is the prototypical invention and exploits the following steps:

• Identify need
• Identify constraints
• Identify candidates to satisfy even part of need
• Determine how far short individual candidates fall

The second type may be thought of as product development within the framework of an existing design. The identification and correction of "bugs" ties invention to continuous improvement. The third type of invention is what NASA (for instance) does regularly, calling it "Spinoffs."

Within an historical context, some inventions have arisen when the societal conditions were ripe: enough money (or demand) available; prerequisite technology in place; no cultural taboos to lead to opposition. Some think that invention is therefore deterministic: at the right time, someone will do the inventing -- it doesn't matter who. Once in a while, however, someone arises who has either the creative genius or the capital, or both, to pick up on ideas that have been missed or were incomplete.

Here are types of improvements that can be made in existing products, or ties between families of devices that lead to new inventions:

1. A joining of existing devices.
2. The repetition, relocation or replacement of features.
3. Disposable sections.
4. Simple human interfaces.

Design changes focus on durability, portability, simplicity of human interface and performance. Such focus, in purpose and measurable standards, separates invention from Darwinian evolution.

Can the focus be accomplished? Here are considerations:

• With multiple problems, you must know which to attack -- and which to ignore.
• A smooth seam between touching subproblems makes an otherwise tough solution possible.
• Parsing, or functional separation, enables the independent tuning of functions, particularly when those functions are separated in space.

Weber lists the following heuristics (or, learning or problem-solving aids) of invention:

1. Feature addition or deletion.
2. Make-variables: identify parameters that can be varied, and vary them to approach some desired measurable goal.
3. Fine-tuning: rearrange, tweak.
4. Variance-control: by minimizing or spreading range.
5. Interpolation and extrapolation: suggests that intermediate states in development may be useful, and that trends suggest the next step.
6. Inverse: join opposite functions.
7. Complement: join functions used in the same context.
8. Shared-property: join functions that share parts.
9. Emergent functions: watch for capabilities not present in previous versions of device; consider ability to use multiple functions simultaneously.

Here's a Web site with some interesting thoughts on invention.

Newsgroup misc.creativity is dedicated to the creative process. Those who wish to discuss this process should check with that group for information. That group is very noisy, but there is a FAQ.

Goal/QPC defines the Seven Creativity Tools, developed at Germany's Batelle Institute in the 1970s:

1. Problem Definition. To help in determining the most likely successful perspective on solving a given problem.
2. Brainstorming. The teaming of people of various backgrounds with people familiar with the problem to be solved. The exchange of ideas stimulates thought leading to solutions.
3. Brainwriting. An idea is written on paper and passed along to another member of the team, who adds on paper an idea triggered by the previous, and so on until all have contributed.
4. Creative Brainstorming. Applying results of separate brainstorms to one another to check for applicability.
5. Word and Picture Association. The use of pictures, words or inventions to generate new ideas.
6. Advanced Analogies. The study of examples of how people have solved similar problems.
7. Morphological Chart. Using parameters of the desired solution, matching mutually exclusive options of each parameter to determine possible solutions.

Goal/QPC claims that their adoption of these creativity tools as a consultation product has led them into teaching systematic innovation. These are, however, only tools, and should be treated as such. The most important consideration in an innovative organization isn't the tools -- it's the environment. Adams (in The Dilbert Principle) cites his own axiom:

"...creativity is allowing yourself to make mistakes; art is knowing which ones to keep."

What Adams suggests managers must do to encourage creativity is as follows:

• let employees dress/decorate workspaces/format internal memos any way they want
• eliminate quality teams and employee suggestion plans (see the FAQ on Engineers and Quality for more on this)
• have a reliable system for internal communication
• make sure employees learn something every day (and don't make them work mandatory overtime to ensure it)
• teach efficiency in "secondary" tasks, such as meetings, memos, forms, hallway conversations and other low priorities

Adams goes under two basic assumptions here: first, that the average person is only mentally productive a couple hours per day; second, that the best way to ensure happy, productive and creative employees is to let them walk away from their work as much as possible.

Revised somewhat to account for non-military applications, and derived from Rich & Janos.

1. The program manager must have complete control over technical, financial, and operational matters.
2. Strong but SMALL project offices must be provided by both the Skunk Works and its customers.
3. Use a SMALL number of GOOD people.
4. Use a SIMPLE drawing system with great flexibility in case of failures.
5. Use a SMALL number of GOOD reports, covering only IMPORTANT work.
6. Have regular, timely budget reviews covering not only actual but also projected costs. Don't surprise the customers.
7. The Skunk Works must be given sole responsibility for negotiations with vendors and subcontractors.
8. Basic inspection responsibility must be left with vendors and subcontractors; the Skunk Works must be responsible for inspection of the integrated system. Inspection procedures must be agreed upon by the Skunk Works and the customer in advance.
9. The Skunk Works must be given sole responsibility for system testing, both in the prototype and advanced stages.
10. Hardware specifications must be agreed upon in advance.
11. Funding must be regular and timely.
12. There must be close cooperation between the Skunk Works and the customer on a day-to-day basis.
13. Access by outsiders must be controlled to a reasonable level.
14. Good performance must be rewarded according to measures other than the number of people supervised, the number of parts fabricated, etc.

Taking a closer look at these points, you can see that it's important to understand up front the definitions of terms such as "good," "simple," "small," "important," "regular," and "timely."

Having decided that you will encourage creativity in your shop, here are some characteristics of a good design, from the perspective of the potential user (per Norman):

• possible functions easily recognizable
• alternatives to primary function (and their possible results) evident
• current status indicated
• "natural" mappings of control to function

These characteristics assume that the object in question is a machine, pretty much. That it has some sort of on/off and function controller or level setting switch or knob. The term "mapping" connotes the path from control to action.

How to accomplish good design:

• design assuming error
• simplify
• standardize
• make everything visible
• make mappings direct
• take advantage of constraints
• use "common sense" (where it exists)

How to design assuming error: recognize the types of error that are known for the thing you're developing:

• apply frequently-taken actions to wrong situation
• take the wrong action with the right intent
• take the wrong action based on the right information
• take the wrong action because it's similar to the right action
• forget what the right action is
• choose the wrong action when more than one is possible

Because some actions are easy to mess up, you have to be careful not to design so that dangerous actions are easy to carry out.

How to make your design better able to accommodate error, once you recognize what error is for your case:

• make it impossible to do some things until certain other things are done first
• make precision unnecessary as long as the right action is taken
• decide whether each aspect of your design must be learned by memorization, by association, or by some built-in explanation

Here is a great website for learning how NOT to do good design: baddesigns.com.

Newsgroups alt.inventors and misc.int-property have patenting as part of their charters. You can find much more information in those groups. Watch especially for threads dealing with "PATNEWS," news updates on patenting and patent law.

Having invented something, you will probably want to patent it, so that you can get it to market. You can file a patent application yourself for as little as US $75. You may also have someone else handle the patent application -- it is becoming increasingly common to have an organization with experience assist inventors with development, licensing and marketing as well. In that case, you must be careful not to overpay for results that are not guaranteed, and not to give away control of your invention.

Here are some Web sites with information on patenting.

• USA Patent & Trademark Office
• patenting and invention
• answers to patenting questions
• USA Federal Trade Commission on invention promotion scams
• Chemical abstracts, patents back to 1971
• USA patent search

The US Patent & TM Office offers abstracts on CD-ROM for $300/yr. (703) 306-2600.

The following is condensed from a pamphlet on patenting from the USA Federal Trade Commission on Invention Promotion Firms, dated January 1994:

Some invention promotion firms may help you get your idea or invention into the marketplace. But be aware, some inventors have paid thousands of dollars to firms that promised to evaluate, develop, patent, and market inventions and got nothing for their money. So be cautious. Your enthusiasm for your idea may make you vulnerable to promoters who make exaggerated claims about the market potential of your invention.

The fact is that some of these invention promotion firms are scams. Don't give them your money without knowing what you get for it. Here is how to identify legitimate firms:

• They can provide hard data regarding inventors they have helped in the past, and they will provide it willingly. They will disclose their success and rejection rates. (A firm that does not reject most of the inventions presented to it is at best over-optimistic.)

• They can support any claim of contacts within the manufacturing sector with names, phone numbers and successful past ventures.

• They will tell you up front exactly and specifically what you would get out of a market evaluation. They will tell you what their own evaluation is based on and how they are qualified to conduct an evaluation, both technically and commercially.

• They will typically rely on royalties from successful licensing of inventions, or at least charge by the specific contacts made, rather than require a retainer to act as your licensing agent.

• They will check your idea against existing patents.

• They will tell you what their costs are, and what they will provide for those costs.

Investigate the company before making any commitments. Call your Better Business Bureau, local consumer protection agency, and Attorney General in your state and the state in which the company is located to learn if they know of any unresolved consumer complaints about the firm.

For More Information

US Patent and Trademark Office     (703) 557-4636 
US Small Business Administration   (800) 827-5722 

United Inventors Association of the United States of America
(UIA-USA)
P.O. Box 50305
St. Louis, Missouri 63105 
(SASE required for list of chapter locations)

National Congress of Inventor Organizations (NCIO)
727 North 600 West
Logan, Utah 84321
(801) 753-0888

Minnesota Inventors Congress
P.O. Box 71
Redwood Falls, Minnesota 56283-0071
(507) 637-2344

Having patented an invention, you may well be on the way to starting a business to develop and market it, or to joining a small business to perform these functions. This section deals with advice for the start-up engineering business and advice for the engineer that wants to hire on with a start-up business.

If you intend to produce and market the product yourself, then pay attention to the following areas:

• Administration. Pay equal attention to all sides of the operation. As a start-up, you are almost certainly too small for your team members to over-specialize.

• Customer contact. Ostentation will not help you in dealing with potential customers; nor will that "three guys working out of their garage" look.

• Design. Not everything worth doing is worth doing well. Design for your customers' needs, not to impress your peers. Stop striving for perfection and ship it.

• Scheduling. What's the earliest date by which you can't prove you won't be finished? That becomes your release/ship date.

• Financing. You'll need a good chunk of capital to get going and stay going long enough to get established. How many sources would depend on what you want/need for equipment, and how hungry people in your geographical area are for a new company in your business area. Most businesses fail due to under-capitalization, and do so within the first three years.

• Building a customer base. If yours is a referral business, then suck up to your local banks, realtors, and lawyers -- these are people who are likely to have work to hand out, and they'll be interested in trying out the new kid on the block. Do not make the mistake of working dirt cheap to build a client base, it does not work.

• Advertising. Nine out of ten people looking for a surveyor (for instance) will look in the phone book. Then a company name like ACME or Aardvark or AAA is a real advantage though a transparent marketing ploy. The other 10 percent will ask their bank/neighbor/lawyer for a name.

• Pricing. You can't have gotten this far without developing a sense of what your services are worth. One suggested rule of thumb is to charge 2.5 times what is paid out for wages. The biggest mistake most engineers and surveyors make is not charging enough for their services. You are doing no one any favors if you are losing money.

...on what is it based? Perceptions, for the most part. Perceptions of quality, strong management, fiscal responsibility, and strategic alliances. Here is a longer list.

• The demonstrated existence of a market for your product.
• The perception of quality. Is this "spin-doctorism?" You must convey to others what you believe yourself about the quality of your product, and the quality of the people on your team.)
• Whether others who have been in the business for a while take you seriously.
• Your ability to convey information -- the right information, in the right amount, and organized the right way -- to people who have money and/or influence.
• Independent estimates of company value. (Since these estimates will themselves be based on perceptions, it's easy to think of them as "vaporware.")
• The perception of strong management.
• The perception of strategic alliances.
• The perception of fiscal responsibility.
• The frequency of new inquiries about your technology.
• An aggressive schedule, with specific milestones.
• The vigorous pursuit of product improvement.
• The potential for new patents and/or derivative technology.
• An opportunity for investors and partners to "kick the tires."

Since you'll need "sophisticated investors" in some cases, you'll have the responsibility of demonstrating most, if not all, of the points given above. Many of them will be expected to appear in your business plan.

It is possible, however, to find a company whose business it is to legitimately do invention marketing. It's more than just getting some paperwork and calling (or pretending to call, in the case of some of the scams) a few larger companies. A full- service invention marketer will be involved in each of the following areas:

• prototyping
• soliciting for license agreements
• soliciting for development funding
• planning for commercialization
• pursuing patents on the entire invention, components, or processes leading up to it

If you decide not to produce and market the product yourself, you may also sell the patent outright, or license it. Before you sell, consult with a patent attorney. Otherwise, consider the following:

There was a wonderful article in the 03/07/96 Machine Design: Rein, B. "Permission Granted." Rein, a (shudder) lawyer :-), went through the options for licensing inventions. To license is to authorize someone to do something they otherwise wouldn't be allowed to do. In this case, it's manufacturing or selling your invention, or both.

Here are the benefits of licensing:

• profit from the technology of others which might otherwise not be available to you
• profit without a substantial capital investment
• avoidance of legal problems involved in foreign jurisdictions
• expanding market geographically

Here is when you want to consider licensing:

• when your technology is useful in several fields
• when your technology is ready to go into new (geographic) markets

Here is what you want to consider in preparing to license your technology:

• what role everybody involved is going to play
• what everybody is going to get out of it
• whether you're going to allow the licensee to make, use, or sell (or some combination)
• what geographical territory is covered
• whether or not the license is exclusive
• how you're to be paid
• under what conditions the license agreement ends
• how agreements and patents are to be enforced
• whether you get to use any related technology the licensee develops

• Get all agreements in writing. Reflect carefully on your requirements, and lay them out. Pick out your minimum requirements, add 20-50% and present them, for negotiating room. And for any change they suggest for less than you ask; ask for something more to make up for that loss (e.g. company car, longer vacation, etc.).

• Make certain that you are offered the technical challenge and autonomy that you are denied in your present job. Or were denied in your last job. Or that you at least get to do something interesting, since if you've been out of work a while you'll take anything. :-)

• You may be offered profit-sharing.
  
  • NEVER give up anything in return for it. Stock options, for instance, require you to lay out money. If you can, get a piece of the gross sales instead. It's easily tracked, based upon shipments.
    
  • Stock in lieu of salary is another form of payment. This can really be advantageous (costs them almost nothing - especially helping cash flow) if they take off, because you collect even if you're not there. Numbers of shares in this case are meaningless -- instead talk percentages.
    
  • If they are extremely solid (like HP, etc. -- not many of those around), 10% value of salary in stock options per year is reasonable. If less solid, 33 1/3% to more than 50%, even 100%, of salary is not at all out of range! (But your mileage may vary.)
    
  • Gauge the extent of stock dilution as a result of more being issued. This of course affects the value of what you get.

• Be prepared. Any downturn in a start-up and the engineering people are the first to go. Make sure your agreement lives beyond your employment, else they will fire you just to cut you out.

• Be prepared to walk!! If it's not the deal that's good for you, you can't take it. They'll not respect you otherwise, so you are at that point backed into a corner.

This section is under construction -- more so even than the rest of this document. Anybody want to contribute? Here's a good place.

Discussions in sci.engr indicate that "rapid prototyping" is defined by the ability of your CAD software to participate in the process. Others have experience, however, in simply developing prototypes in a hurry. That calls for one of the following:

• Parts readily available in-house
• Participation of vendors

The Four Laws of Dealing With Vendors

I	You know the application. The vendor knows the product.
II	The more information you give, the faster the turnaround and the more reliable the price.
III	The closer your choice is to a stock item, the better the service.
IV	The vendor wants you to succeed, knowing that if you do you will pay your invoices for a long time.

First Law. Contacting vendors turns out to be a golden opportunity for the young engineer (any engineer?) to learn not only about the vendor's specific product line, but about the type of product in general. At first, it's likely that the product you are shopping for has more characteristics than you need to know for your application. (If you're looking at a menu, you choose your dinner on the basis of cost and what's in it -- not on the individual "Nutrition Facts.")

In one example, a contributor was looking at load cells, and one load cell manufacturer gave specs on the following: rated output, excitation, accuracy, linearity, hysteresis, repeatability, zero balance, creep, operating and storage temperature, thermal effects, maximum load, bridge resistance, deflection, materials used in construction (both of structural and electrical parts), height, weight, bore, compatible meters, rated capacity, and cost. The contributor's interests were rated capacity, geometry and cost. That other stuff is listed in the catalog for a reason, and no doubt that reason is related to some customer's needs. Knowing what others care about in their purchases may guide you in a purchase down the road, when your own requirements are more severe.

Talk to vendors on the phone. You can nearly always learn about how their thing works, and what you should be looking for in a purchase. In another example, a contributor was looking for electric winches for pulling wire rope. The first vendor explained why the winch's drum size is closely related to its rated pull and power draw. The second vendor explained why the rated pull of the winch is only about one-sixth to one-fifth the rated strength of the cable. The third vendor explained how a winch can backdrive, what the consequences are of the backdrive and how to avoid potential problems in winch selection. The contributor also learned also about grooved drums, mounting technique and noise.

Use vendor catalogs. They can contain a wealth of knowledge of product fundamentals. The Editor has not seen a large number of them, but he recommends as examples Omega Engineering catalogs because of the info they contain about fundamentals of sensing devices; and the National Instruments "Instrumentation Reference and Catalogue" for data acquisition fundamentals. Best of all, such catalogs are to be had for FREE.

Design spec sheets for your application. In the winch example, info sheets listing requirements on winches, and on wire rope, were faxed to dozens of companies in each of the two areas. This task quickly eliminated three-quarters of the possible companies from contention. Which leads to the Second Law.

Second Law. Why will you find so many companies in the area you are looking for?

• You may not be sure who to contact up front, unless you have done this before.
• The trade literature doesn't list vendors according to your specific needs. For instance, the Thomas Register lists over a hundred companies under "Winches: Electric." Although the Thomas Register has a search engine at its Web site, which will reduce the number somewhat, the reduction will not specialize your search any further. No search engine is smart enough to recognize your requirements.

At the Chinese (for example) restaurant, if you tell the person taking your order up front that you require your meal to be spicy, contain no shellfish and no MSG and no baby corn, then that helps you get a meal to your liking with no long wait and no surprise cost. In the same way, the faxed requirement sheets may keep you from having to call each company individually -- and no doubt from playing "phone tag" with some. You may even locate a company with something on the shelf that meets your needs. Which leads to the Third Law.

Third Law. If you choose something off the menu, it may cost more, it may take longer, and they may not even have it at the restaurant. The same contributor encountered this when trying to purchase conduit for the wire rope. The wire rope was aircraft cable -- which has a rated diameter with a tolerance of ten percent -- the "rated diameter" is actually a "minimum." They wanted conduit that would permit the cable to pass without slop, but would also accommodate the relatively huge tolerance in cable diameter. (This is expensive.)

Buy multiple related items from the same supplier when possible. You may (in this example) find you are better off buying both cable and conduit from the same supplier -- even if it means letting a spool of aircraft cable go to waste. If a vendor doesn't have what you want, that vendor would either not bother to respond or send a polite "decline to bid."

Let the vendor be the expert on the vendor's product. Our relationships with vendors can make us more productive, by actually eliminating the need for our project engineers themselves to be experts on the type of product the vendor sells. Follow as much as is practical the vendor's recommendation.

Fourth Law. Why should it be that a vendor (whom you have never met) would be interested in your success? Because if you succeed, and if you attribute your success (in part) to that vendor, you will do business with that vendor for a long time. A satisfied customer is worth more than Gold-Pressed Latinum.

If you pay your bills early, you can get a line of credit later. If your bills are small and your product especially promising, you may get that line of credit early. Don't hesitate to ask for it -- but be careful how you approach it. You must remember the First Law of New Technology: no-one wants to be the first to buy it. So in approaching vendors, as you would in approaching potential investors, you must concentrate on the promise (and the market niche) of the technology, rather than on its developmental aspect.

Don't hesitate to let the vendor know how you're doing. Establish a personal contact at the vendor's shop. (Most will want to give you one. That person will probably be referred to as an "account representative," or something like that.) If you have a personable relationship with that representative, then you have an advocate at the vendor's shop. Comes in handy.

Characteristics of "Good" Vendors

These are in alphabetical order. You can decide which are the most important for your needs.

Experience	A good vendor knows "the business." They can tell you if you're going in the right direction. They can provide alternatives which may even fall outside the product line.
Flexibility	Most vendors "specialize in large runs." Well, of course they're going to specialize in orders likely to bring in the most money. But a good vendor, even if unable to accommodate a very small run, will come up with an alternative: perhaps a stock item that's close and they'll let you have it FREE or on a long-term "loan."
Knowledge	A good vendor knows the product. Not just what it says in the catalog, but actual applications. The vendor should be able to go out and look at an installation or drawing and make suggestions with merit quickly. Newbies will make a call to someone at their firm for help.
Persistence	A good vendor will call on you even though you don't order anything for some time. They know if they are there now, you'll be more inclined to think of them later.
Timeliness	A good vendor will have the information within a day or so, even if you don't need it that soon. Some will have a sample in stock for you to try out for FREE.

Characteristics of "Bad" Vendors

Poor Service	Calls not returned; problems not taken care of. Why should such a company be rewarded for poor service? On the basis of its name? Or because "we've always done business with them?"
No Professionalism	Some vendors give out business cards with nicknames; some give out silly and inappropriate toys; some make racist, sexist, or simply stupid comments; some argue with you about your project, or tell you it won't work. Why should you pay someone to make you put up with all this?

How to Find Vendors

The Thomas Register, or Vendor Directories. Since the Thomas Register is on the Web (and on CD-ROM), you might be able to get by without, you know, actually buying that couple hundred pounds of catalogs they put out in print. (A small company is always looking for the main chance.) But many, many vendors are getting on the Web, and many of them will no doubt be found at www.(insert company name here).com.

Trade magazines. These also prove very useful guides to locating vendors. Machine Design, for instance, annually publishes a "Product Locator." The trade magazines are often kept afloat by their advertisers, and thus you can get subscriptions to some of them for FREE. Be careful with your request, however: Machine Design (for instance) is one publication that routinely rejects requests for free subscriptions from very small companies. Some from small start-up companies lie about how many people they work with to get these subscriptions.

The vendors themselves. One participant in this discussion saves time by demanding that what he buys be second-sourced. "Ask your vendor who supplies the same thing? And the list gets self-generated. Next, ask each one why you should buy their product and not from their competitor (be specific) You'll be amazed at the 'inside' information you'll get. Also, ask for a list of satisfied customers and talk to them, after all you don't hire a potential employee without checking their references why would you buy a product from someone without checking theirs?" The reasons this idea is so good are as follows:

1. Many of our employers would have the multiple-sources requirement on paper anyway.
2. Each vendor runs into the same competition over and over.
3. Some of them send business to each other occasionally as a matter of course -- when the competition has a peculiar specialty the vendor in question doesn't have.

Manufacturer's reps. They will sometimes represent several companies with similar products, saving you the hassle of dealing with multiple companies. One contributor claims to be able to limit a product search to 2 or 3 reps, rather than 10 or more individual companies. The reps are good at giving comparisons among the companies, and they are usually very knowledgable about their field.