Follow @jack_ganssle


The logo for The Embedded Muse For novel ideas about building embedded systems (both hardware and firmware), join the 27,000+ engineers who subscribe to The Embedded Muse, a free biweekly newsletter. The Muse has no hype, no vendor PR. It takes just a few seconds (just enter your email, which is shared with absolutely no one) to subscribe.

By Jack Ganssle

Published in Embedded Systems Design, July 30, 2010

Protecting Your IP

Thomas Edison became a folk hero in no small part due to the perceived embodiment of him as a typical American, bootstrapping himself from obscurity to fame and wealth via a series of inventions. The notion of the backyard mechanic/tinkerer is deeply ingrained in the US story, as it should be. We are a nation of problem-solvers, our practical Yankee ingenuity having morphed from improving plows to putting billions of transistors on a fleck of silicon.

Our Constitution protects the rights of inventors. Article I, section 8 grants Congress the power "to promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries." This one phrase spawned the 10,000 employee Patent and Trademark Office (USPTO), which is a part of the Department of Commerce.

Interestingly, another Thomas, the inventor at Monticello, was opposed to any form of monopoly, including patents. At one point he wanted an addition to the Bill of Rights that would have prohibited all monopolies.

The US system is hardly unique. It draws on the 1623 Statute of Monopolies enacted by Parliament in England, which in turn stemmed from traditions in Renaissance Italy and ancient Greece. Today, the European Patent Office handles patents from 32 member states. That organization, coupled with the USPTO and Japan's Patent Office are called The Trilateral Offices as they process the bulk of patents around the world. China, though, is not far behind. Their growth in applications far outpaces such growth here.

In the US, half a million applications are filed a year. That's nearly five times the number in 1980. But only 38% were granted - down from 59% thirty years ago (source: http://www.uspto.gov/web/offices/ac/ido/oeip/taf/us_stat.pdf). Are ideas getting worse, or are the examiners more demanding?

A quarter of US patent applications are from California inventors. Half of US grants are, amazingly, to people in other countries. 38% of those go to Japan, 10% each to Germany and South Korea, and 8% to Taiwan. Trinidad and Tobago and Vietnam each received two US patents last year (http://www.uspto.gov/web/offices/ac/ido/oeip/taf/st_co_09.htm).

In the last two and a third centuries about 7 million US patents have been granted, including number 6469 to one A. Lincoln in 1849, who later went on to much greater fame. It took just under a century to get to patent number 100,000 (for a horse sun bonnet), another 41 years to reach number 1,000,000, and number 2,000,000 was awarded just 24 years later. Both numbers 1,000,000 and 2,000,000 were for tires, reflecting the growing importance of the auto in our society in the first half of the twentieth century.

Patents come in a handful of flavors, but few are other than utility patents, which protect an invention or an improvement to an existing invention. A utility patent stifles competition for 20 years from the date of filing (not the date of the grant). Figure around two years to go from filing to an issued patent.

Patents are typically written using a quite stilted form of the English language. They have several sections, but the most important are the claims and the preferred embodiment (also known as the specification).

The claims are the real meat of the patent and that's where you'll find what the inventor is, well, claiming as his unique contribution. Any number may exist, though the number does effect the cost of obtaining the patent. An example claim (from patent 6,000,000) is:

1. A method of sharing information on a first computer system and a second computer system, said method comprising:

connecting said first computer system to said second computer system with a data communications link;

providing a library of functions in said second computer system for accessing information on said first computer system;

creating a conduit program database, said conduit program database for storing a list of conduit programs that may be executed, registering a first conduit program by placing an identifier for said first conduit program in said conduit program database, said first conduit program comprising a computer program on said second computer system for performing a specific data transfer task;

successively executing a set of conduit programs identified within said conduit program database from a manager program, each of said conduit programs accessing said library of functions for communicating with said first computer system.

Each clause is called a "limitation," which further refines the scope of the invention.

The preferred embodiment describes the invention in detail, generally with drawings. Its role is to describe the device at least enough so "one skilled in the art" can produce a copy without "undue experimentation," which is a vague-sounding phrase mediated by lots of well-understood law. Inventors make a deal with the government: a patent gives one a sort of monopoly for a time, but to further the development of new ideas the inventor must bare all to the public. The preferred embodiment does that, but realize that there are no secrets.

Costs are high. Though for an individual or small business the USPTO only collects a few hundred dollars in initial fees, navigating the waters is complex business so most inventors hire an attorney. That's probably smart because the right legal mind will help form claims that are broad enough to be useful, narrow enough to be patentable, and worded in such a way that will aid in the future should the patent be infringed. But lawyers don't come cheap: even a simple patent could cost $5k, with more complex ones, or ones requiring a lot of give and take with the examiner, costing over $15k. And there's no guarantee that the patent will be approved.

Then there are the maintenance fees. 3, 7, and 11 years after the patent is issued even an individual must pay the Feds to keep it in force. That's currently about $4k total for the three installments, with larger entities paying roughly twice that.

Litigation

Considering the cost, hassle, and the lack of any sort of trade secret protection, why bother patenting anything? Beyond being great resume fodder I see two main advantages: first, the existence of the patent may prevent others from reusing the idea. Second, if a competitor knowingly or otherwise makes use of the concepts then one can sue for damages and to keep the infringing device off the market.

Litigation keeps the JD grads in Gucci shoes. The average cost to prosecute or defend a patent suit is about $2 million (source: http://www.jmripl.com/Publications/Vol4/Issue1/ronspies.pdf). And it can cost a lot more.

Going to court is also slow. Robert Kearns, for instance, invented the intermittent windshield wiper in 1964. He offered it to the Big Three but was spurned. They started selling such wipers just five years later. He sued Ford in 1978 and Chrysler four years later. The first case was at suit for 12 years; the second for 10. The financial resources and patience required can beggar the imagination.

A device can infringe on one or more of a patent's claims; each stands alone as a sort of invention in itself. The patent owner decides which claims are worth litigating, though that may change throughout the case due to negotiations with the other side or for other reasons.

When a claim consists of several limitations an accused widget infringes the claim if it embodies each and every limitation in that claim. So if a claim consists of a limitation that the device makes red paint and another that it makes blue paint, the widget has to make both red and blue to be found infringing.

To make the legal process more interesting, torturous and expensive, the court may be asked by the warring parties to "construe" words or phrases in the claims. For instance, in the claim above one limitation reads "connecting said first computer system to said second computer system with a data communications link." What does "data communications link" mean? Does that include smoke signals? A telegraph? Or just an OC-48 line? If asked, the court may define it and then all of the debate must take place within the framework of that definition. The preferred embodiment is used as a guide to the meaning of the terms.

And to think I used to laugh about Bill Clinton's struggles with "is."

There are no Perry Mason moments when litigating patents (the eponymous show ended in 1966, so for younger readers, Perry Mason was first a book and then TV program staring a lawyer who inevitably brought his cases to a successful conclusion by revealing some startling piece of evidence during the trial.) Long before trial starts each side has all of the incriminating material, and often much more. "Discovery" is a process where parties can demand documents and other material from the other side. That can run to the millions of pages and often includes electronic records. So if you're working on patentable equipment, be careful with your emails.

That's probably good advice for climate scientists, too.

The party holding the patent must prove to the jury that the accused device infringes the stated claims. That might be as simple as a smack-the-head "duh!" or could involve deep analysis of the device by the plaintiff's experts. They, then, testify in court in front of the jury.

In the cases I'm aware of the patent-holders also have to show that they sell a product that uses the claims. The notion here is one of damages: if an infringing device goes to market, presumably that will harm the patent-holder.

There are basically two defenses against a charge of infringement. First, one can try and prove that the patent isn't valid. That's tough; the law presumes that the USPTO does a good job. But it happens when it can be shown, for instance, that the invention would have been brain-dead obvious when it was patented.

The second defense is that the accused device just doesn't infringe. Again, experts and other witnesses will provide testimony to sway the jury.

Ultimately the jury reaches a verdict. One party goes home and licks its wounds while the other breaks out the bubbly.

Some readers may find that the litigation I've described is awful, disgusting, a perversion of our way of life. We hear a lot about greedy ambulance chasers. I'm sure there's some of that. But in court I'm overwhelmed by the magnificence of the process. The two parties and counsel are sitting politely waiting for their arguments to be heard. The judge is there to insure fairness. A court reporter takes down every word so there will be no dissent on what was said by whom. Evidence is presented and discussed. A jury, hopefully of our peers, listens to that evidence and debates the truth. There's no lynching, no six-shooters, and logic is (generally) used rather than raw emotions.

Is it a perfect system? Of course not. But the idea of aggrieved parties peaceably bringing their differences to court and then (by and large) accepting the jury's verdict is enshrined in our Constitution and way of life.

(The fine print: I'm just an engineer - by all means it pays to get a lawyer if you're engaged in any fashion with the justice system.)