Do you need a working prototype of your invention to file a patent?

During the 1970s, I was 16 and became interested in inventing and anything involving solar energy. Solar energy was quite the rage after the 1973 oil embargo and fuel crisis of that era.

I enjoyed building things and basically thought inventing something meant building a working prototype of it, so I built a parabolic solar reflector that had a glass tube filled with black-dyed water/liquid that would flow through the focal point (hot spot) of the reflector. On a sunny day if I didn’t flow my liquid fast enough, the black liquid would boil and crack my tubing. I naively thought all I needed to do to get a patent was to send some pictures of my invention prototype to the U.S. Patent and Trademark Office (USPTO), along with some cursory paperwork.

I learned that no prototype or model of an invention was required—and the required paperwork was a bit more involved than cursory. No working model required? How could a patent examiner, or for that matter anyone, be sure an invention worked?

The era of the patent model

I discovered I wasn’t the only one who thought this way—it included the USPTO, for an extended period. They required models with all submitted patents for nearly 100 years—from 1790 to 1880. Congress abolished the legal requirement in 1870, but the USPTO kept the requirement until 1880. It was founded in 1790, so this requirement was there at its inception.¹

These models were small-scale three-dimensional (3D) working representations of inventions. They could not be larger than 12 × 12 × 12 in. Patent models enabled inventors to show the function and elements of their invention without having to describe them, and eliminated a need for thorough and complex diagrams. The models enabled easy comparison with existing inventions and a certain level of physical validation that the invention worked as claimed. All of this was in line with my view of how a patent application for an invention should work.

The ending of this requirement for a physical prototype by the USPTO was preceded by two major fires in 1836, and another in 1877, which destroyed 76,000 models.² These fires highlighted how much space the models took up. A plot above shows logarithmically how the submission of patent applications would grow within the U.S. after this period (~1880). During the first year the patent office opened, 765 applications were submitted. During 2021, more than 650,000 patent applications were submitted. If models were still a requirement, there would be need for a building the size of the U.S. Capitol Rotunda (~13,000 ft³) to be built every two years to store all of these models if they were piled one upon another.

Patent model museums

In 1926, the USPTO decided to rid themselves of most of these models. They kept ~15,000 and auctioned off the remaining models.

You can still see many of these 140+ year-old models in both private and public museum venues, including the following:

• Rothchild/Hagley Museum & Library – Wilmington, Delaware³
• Henry Ford Museum - Dearborn, Michigan⁴
• Smithsonian American Art Museum (Old Patent Office) – Washington, DC⁵
• Smithsonian National Museum of American History – Washington, DC⁶
• Susan M. Eichelberg Glendening Patent Model Collection – Cornwall on Hudson, New York⁷
• Harvard Science Center (2nd floor) – Cambridge, Massachusetts⁸

The patent model became a casualty of the explosive nature of innovation within the U.S., and so was born the concept and requirement for “constructive enablement” within the patent application.

Constructive enablement: A descriptive patent model construct

The requirements for a U.S. patent require both a written description and enablement. These might appear to be the same thing, but the USPTO has an entire section (2164) of their Manual of Patent Examining Procedure (MPEP) entitled “The Enablement Requirement.” These requirements are an expansion of the 35 U.S.C. 112, which is the law that describes the core descriptive portion of a patent which is the “Specification.” This is the section of a patent where constructive enablement should be provided by an inventor.

This U.S. code of law, which is the genesis of modern U.S. patent law and harkens back to 1952, and contains the following phrase: “The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.”⁹

This is the enablement clause. It tells us that although a model is not required, the patent must enable the invention to be built by reasonably qualified others. As with most things within the legal realm, this clause has been litigated many times to clarify what it means for an inventor. My desire to better understand what was required to provide “constructive reduction to practice” or “constructive enablement” led me to read opinions in these matters. I found one relatively recent case quite helpful. It was a Federal Circuit case called Ariad Pharms v. Eli Lilly & Co., from 2010. This decision reaffirmed that 35 U.S.C. § 112, noted above, contains both a written description requirement and an enablement requirement—and they’re separate and distinct.

Of particular interest here are a few “broad principles” articulated about “constructive enablement” or the modern words-only version of the patent model. They are summarized by the end noted source¹⁰ as follows:

• The written description requirement does not demand either examples or an actual reduction to practice.
• A constructive reduction to practice that in a definite way identifies the claimed invention may be sufficient.
• Actual “possession” or reduction to practice outside of the specification is not sufficient—the specification itself must demonstrate possession.
• A description that merely renders the invention obvious does not satisfy the requirement.

No actual prototype or model is required to file a patent. A constructive (written) reduction to practice suffices if it demonstrates possession of how to make the invention. This possession must be explicit and not just implicitly show the invention to be obvious and left to the reader’s imagination as to its construction.

It’s interesting to note the USPTO left the door open for requiring physical models. In 35 U.S.C. 114, we find the following: “The Director may require the applicant to furnish a model of convenient size to exhibit advantageously the several parts of his invention.” Some inventions are so fantastical and deftly described that it’s difficult to determine whether their provided constructive reduction to practice is real or not. Perpetual motion machines or precious metal transmuting processes are among a few that elicit this requirement from the Patent Office to this day.

As an inventor, unless you are confident that you can build and demonstrate your invention physically, then you will probably find it difficult to provide adequate constructive enablement. In this case, building a working prototype helps you understand any possible issues. You may discover issues with functionality that you did not perceive, and the resolution of those issues requires critical elements in the invention that would not otherwise be in your patent submission. Although the USPTO does not require physical models for patent submission, building and testing a succession of prototypes for your invention remains the most powerful tool available to an inventor. Build, test, resolve, and build again is the cycle of effective invention and, ultimately, innovation.

Why this matters

The modern version of the patent model is called constructive enablement. Constructive is a legal euphemism for “written.”

An invention can be quite simple if you know the right or critical problem that needs to be solved for your customer. An apparent and straightforward solution that is also inventive might be understood immediately. In these instances, the true intellectual property (IP) is the realization of the commercially valuable problem. It’s advisable to write an invention disclosure immediately and get a patent filed prior to the delay of any physical model being built in this case. It might seem like jumping the gun a bit, but opportunity for HP to own valuable IP is fleeting. This is especially true since the U.S., in 2013, implemented the America Invents Act, which gave priority to the first to file over the first to invent. This type of strategic early filing is just good practice.

Some personal examples of when pulling the trigger early on a patent filing for these reasons include the following from the fields of data storage and machine vision. While working on super floppy disk cartridges (1990’s ZIP Disk and Floptical Disk) where the rotation speeds were an order of magnitude faster (~3,000 RPM) than the legacy floppy disk (300 RPM) and the use of lubricious fabric wiper materials such as Teflon fiber-based Gore-Tex in the cartridges was of obvious utility for drive motor torque reduction and reduced magnetic media wear. We immediately filed a patent.

Another more recent, purely constructive-enablement-only HP patent filing was for the use of 3D stereoscopic camera system in an electro-optical (EO) digital pen called the Cheetah EO Digital Pen. The problem was the reduction of the computational load of decoding an infrared visible location-encoded dot pattern with the associate parallax distortion of the pattern throughout the various random tilt orientations of the pen while the user scribed. The invention provided acceleration of the computational aspects of this problem, of better than an order of magnitude, and therefore a dual-camera or stereo-vision approach was obvious, and a patent was filed prior to actual physical prototype enablement.

These examples illustrate the need, in some cases, to move rapidly to filing a patent with only constructive enablement prior to a full and functional model of the invention being completed. This, in our first-to-file priority patent system, is strategically smart patent practice. But being able to file a patent with constructive enablement alone doesn’t reduce the importance and criticality of cycling through as many prototypes of an inventive concept as rapidly as possible. This is core to being an effective and innovative company.

Once a patent is filed, we need to remember that we can then take the risk out of the important exercise of sharing the prototype version(s) of the invention/product with potential customers. Even with confidentiality agreements in place, there is risk to the IP being conferred to competitors and losing our patent priority. On the other hand, by not sharing prototypes of a nascent product to future potential customers we lose critical feedback and a better understanding of product essential features we need for it to succeed in the marketplace. The significant value of these customer-forward prototypes is explored in depth over 250+ pages in the book The Innovator’s Method, which is worth any innovator’s time.

Although it no longer requires an actual model or prototype to file for a U.S. patent, innovation does. Become masters of both! It is within all of us to innovate. Today is a good day to file an invention disclosure. Just do it!

REFERENCES

1. See www.invent.org/blog/intellectual-property/patent-model.

2. See en.wikipedia.org/wiki/1877_U.S._Patent_Office_fire.

3. See www.atlasobscura.com/places/rothschild-patent-museum. 

4. See www.thehenryford.org/collections-and-research/digital-collections/expert-sets/10903.

5. See americanart.si.edu/blog-post/560/in-this-case-patent-models.

6. See www.patentmodelmuseum.org.

7. See www.patentmodelmuseum.org.

8. See news.harvard.edu/gazette/story/2009/07/history-on-a-small-scale-2.

9. See www.uspto.gov/web/offices/pac/mpep/s2164.html. 

10. See www.dorsey.com.