The Lemelson Situation

Among the issues facing the machine vision industry today are those created by what has been referred to as 'the Lemelson situation.'

Who is Jerome Lemelson, and what does he have to do with machine vision?  A couple decades ago, when the first working vision systems were being built and tested, if you had addressed that question to the engineers who pioneered the field, you might have gotten some blank stares.  Not so today.

Jerome Lemelson is the named inventor on more than 500 patents.  The subjects of his patents run from toys to bar coding to machine vision systems.  It is claimed (though not without considerable dispute) that Lemelson's inventions can be found in VCRs, camcorders, Walkman® audio devices, cordless telephones, fax machines, data and word processing systems, and industrial robots.

Add machine vision to the list.  Mr. Lemelson is named as the inventor on twelve patents that his lawyers say cover machine vision systems.  Beginning in the early 1990s, Mr. Lemelson began a massive campaign to enforce and license his 'machine vision' patents, and that campaign has been continued by his lawyers and his foundation since his death in 1997.  Although the body of these patents was written in 1954, the Lemelson legal team claims that they cover modern machine vision systems developed decades later.

By all accounts, the licensing and enforcement campaign has been successful.  The list of licensees who have paid up to settle accusations of patent infringement is large, and continues to grow.  This article will provide some background information for the many people throughout the machine vision industry asking, 'What's the deal with this Lemelson situation?'

Lemelson's Machine Vision Patents

Lemelson's patents on machine vision all stem from two patent applications, filed in 1954 and 1956.  The first patent issued in 1963.  Lemelson kept the underlying applications pending in the Patent Office by filing a series of 'continuation' applications, a procedure which allows the inventor to add more claims, and in some cases more description, to the patent.  Thus, Lemelson patents have issued throughout the 1970-1994 time frame, while claiming an invention priority date back to the mid-1950s.  At present, there are hundreds of machine vision claims among 12 patents.

Despite the issuance of a dozen patents, it is far from clear what actual contribution, if any, Mr. Lemelson's work made to the field of machine vision.  The particular machine vision system described in Lemelson's patents was never built as a practical machine in the real world.  Ask the engineers who built the first working vision systems to name the giants on whose shoulders they stood, and see if the name Lemelson is mentioned.
 
Since the beginning of Lemelson's patent campaign, patent lawyers have raised serious doubts regarding both the validity and scope of the claims of the Lemelson machine vision patents.  There are a number of legal deficiencies that, if proven, would render the patents invalid.  Among these is the theory that the patent body fails to describe a workable invention.  The Patent Law requires that a patent must describe how to make and use the invention.  The description much be complete and detailed enough so that an ordinary engineer could pick up the patent, and working from that, build a working system using the technology and know-how available at the time.  The theory regarding the Lemelson machine vision patents is that no one using 1954 knowledge and technology could have built a machine vision system capable of functioning as described in Lemelson's patents.  In addition to this theory, there are other legal defenses that could support a very narrow reading of the claims so that, even if they were considered valid, they could not reasonably be interpreted as covering modern-day machine vision systems. 

So far, however, those legal battles have never been fought to their conclusion.  Why?   Some might say it's because the patents are strong, but this appears to be the minority opinion, and in any event there are other explanations.  The prevailing view is that it's because the license fees offered by the Lemelson Foundation are low compared to the cost of challenging the patents.  In other words, most companies have apparently decided that fighting doesn't make economic sense -- even if they believe they don't infringe or the patents are invalid.  In the few situations where companies do stand their ground and fight, similar economic considerations would seem to apply for the Foundation.  In those situations, it looks like the sensible thing for the Foundation to do would be to offer a very low settlement rather than fight to the end and risk having the patents declared invalid.  Better to let a few companies off cheap than risk killing the goose that lays the golden eggs.  So the patents remain largely untested.

Lemelson's Enforcement Activities and Strategy

Since the early 1990s, Lemelson and his Foundation have threatened patent infringement suits against hundreds of companies under the machine vision and other patents, has granted licenses under the patents to hundreds of companies, and has sued hundreds of others.  The Foundation currently has five lawsuits pending against close to 500 companies.

The basic strategy for Lemelson has been to avoid the machine vision systems manufacturers, and instead target the users for its assertions of the Lemelson patents.  It makes sense.  Chances are that an individual user has less at stake than the manufacturer, and is less able to defend against the claims because it lacks expertise in the technology.  Thus, the economics driving the decision to take a license instead of defend are stronger for a user of the technology than a manufacturer.

The Cognex Suit

Cognex Corporation is a leading manufacturer of machine vision systems.  In 1998, Cognex brought suit against the Lemelson Foundation.  In that suit, Cognex seeks a declaratory judgment that the Lemelson machine vision patents are invalid, unenforceable, and not infringed by any Cognex machine vision systems.  The suit is pending in federal district court in Reno, Nevada.

Market Intelligence News & Insights: 

How Will Apple’s $500 Billion Investment for US Manufacturing, Education, and AI Play Out?

On February 24^th, Apple announced a “commitment… to spend and invest more than $500 billion in the U.S. over the next four years”, with plans for new manufacturing plants, enhanced support for manufacturing partners via their Advanced Manufacturing Fund, and a myriad of other initiatives. As part of this plan, Apple projects 20,000 new hires, with roles focused on R&D, silicon engineering, software development, and AI/ML engineering.

Read More

Without assessing Cognex's chances of succeeding in any part of this endeavor, several factors bear mentioning.  First, this is an important case.  A Cognex victory could bring the Lemelson patent licensing program to a halt.  Second, there are viable theories that the Lemelson machine vision patents are invalid and unenforceable.  Yet, for the most part, in all the litigation over these patents, many of these theories have yet to be considered on the merits by a court.  That is cause for some optimism.  The third point, however, is that a complete defense victory against all 200-plus patent claims is an uphill battle based on the sheer number of claims alone.

What would a Cognex victory mean?  A Cognex victory could mean a number of things to the industry, because it could take a number of different forms.  If Cognex were to succeed in invalidating all claims of all the Lemelson machine vision patents, it would remove the threat from the industry completely.  On the other hand, a victory on a narrower basis, such as non-infringement, might not be as far-reaching.  Either way, it's safe to say that a Cognex victory would likely lead a lot of companies to rethink their negotiating strategy with the Foundation.

When will we know the outcome?  Cognex is predicting that both sides will be ready for  trial by September 2001.  However, motions are presently before the trial court that could give Cognex a victory before then, without a trial.  If the case did go all the way to trial, and Cognex were to win, that still wouldn't be the end of the story.  The Foundation could appeal -- which could take a couple years.  Of course, during the pendency of any such appeal, it's doubtful we would see much licensing activity.

It is still early in the game, and no decisive blow has been struck by either side.  Still, Cognex has achieved some positive results.  Shortly after the filing of the Cognex complaint, a group of bar code manufacturers led by Symbol Technologies filed a similar suit.  Those suits have been consolidated into one proceeding for the pre-trial discovery phase.  The Judge has not ordered consolidation of the trial.

The Judge has ruled on a number of motions; here are some:

• Denied the Foundation's motion to stay the action pending outcome of the Lucent and Intel cases in Phoenix.  In fact, just the reverse happened -- those cases were stayed in view of the Cognex case, over the Foundation's objection.  Those cases are now proceeding on one semiconductor patent despite the Foundation's request to also stay that part in view of Cognex.
• Denied the Foundation's motion to transfer the action to Phoenix.
• Denied the Foundation's motion to transfer the action to Las Vegas.
• Denied the Foundation's motion to dismiss the action for lack of subject matter jurisdiction.
• Granted the Foundation's motion to dismiss Count IV of the complaint, which alleged that the patents are unenforceable because of so-called 'prosecution history laches.'  This is basically an argument that Lemelson should not be allowed to enforce his 'submarine' patents after having kept the applications secret for decades while everybody else actually developed the technology in the real world.  This theory was tried and rejected in an earlier case brought against the Big Three automakers.

The dismissal of Cognex's prosecution history laches defense might only be a temporary setback.  Cognex is seeking to reinstate and pursue this defense.  Recently, Cognex sought and was granted an immediate appeal of the trial court's decision to dismiss this count, and the Federal Circuit Court of Appeals agreed to consider it.  This is a significant development, given the Federal Circuit's traditional reluctance to taking up single-issue appeals while part of a case is still proceeding in the trial court.

Meanwhile back in Reno, the appeal has not slowed down the proceedings in the trial court.  There, the case is progressing through the pre-trial discovery phase, and, as mentioned, Cognex has filed a motion seeking a summary judgment on its remaining claims.  That motion, which could dispose of the case short of trial, is pending and no date for a decision has been set.

The Cognex Letters

In letters sent to a number of companies who have been sued by the Foundation, Cognex has been outspoken about its prospects for winning its suit.  The Cognex letter states that Cognex is determined to stay the course and bring the matter to trial, and is confident that it will prevail and render the Lemelson machine vision patents useless.  It states that time is running out on the Foundation, and predicts that resolution of the latest round of lawsuits by the Foundation might have to await the outcome of the Cognex case.

Conclusion -- Issues for the Industry

The Cognex suit holds out hope that the industry-wide threat of the Lemelson patents could be eradicated within a few years.  In the meantime, many companies facing the Lemelson situation are watching -- and wondering what to do about it today.  Some may decide they have good reason to call a halt to further negotiations with the Lemelson Foundation, and await the outcome of the Cognex suit before finalizing any settlement.

Much is riding on the Cognex suit, and the ultimate outcome is far from clear.  One thing that is clear, however, is that Jerome Lemelson will be remembered by the industry as a significant figure in the history of machine vision.  What that significance is, I leave to you.

.