This is the second half of my interview with John Ferrell, a longtime Silicon Valley patent attorney and angel investor, about intellectual property protection for software. (Read the first half here.)
Ferrell has advised early stage technology companies including Apple, Adobe, Autodesk, Intuit, Oracle, Sega, Facebook, and Sony Computer Entertainment. Full disclosure: I’ve known him for nearly 20 years. In the early 2000s, I hired his firm Carr & Ferrell to help me obtain and defend a portfolio of patents related to a packaging technology.
Since the requirements concerning patentable subject matter related to software vary from country to country, how do you go about obtaining IP internationally?
Over the past 20 years the patent systems of the world have become increasingly harmonized, certainly among the commercially important countries. A patent that works in one jurisdiction can generally be shoehorned to fit into the others.
One thing I try to keep in mind is that tractors and fertilizers are patentable nearly everywhere — they are easy to understand, and the usefulness of each is clear and undeniable.
Whether it’s about an Android app or a new database program, the more I can make a patent claim look like a farm tractor or a pile of fertilizer (in a good way), the easier time I’m going to have at the foreign patent office.
Thinking about a simple tractor, it’s a collection of well-known parts that interact with each other, and function together to perform useful work. For example, there is motor, connected to a drive shaft that turns wheels, that pulls a mechanical disking trailer, for use in tilling farmland prior to planting crops. Four pretty simple parts that if uniquely combined would certainly be patentable.
Similarly, an example of a patentable software program might be an automated music application that helps us find radio stations currently playing songs similar to a song we are just now listening to our smartphone. The music app might contain a signature engine that is capable of generating a signature or short mathematical representation of the song being listened to, a communication protocol for comparing the signature against a database of signatures and for producing a just-like-it list of related song signatures, and a search engine for scanning a directory of known radio stations on the internet for stations playing songs contained on the just-like-it list.
There are three pretty simple elements to this app, but because we have described it as a tractor-like machine, this software program should be patentable almost anywhere.
How important is combining software with hardware in terms of obtaining patent protection?
The three requirements for patentability demand that the invention be new, useful and nonobvious. There are a few subject matter exceptions to what is patentable, for example items contrary to the public good, but there is nothing under U.S. patent law that says software itself is not patentable.
Clearly we need a physical processor to execute the software, but beyond that, there are some software inventions that clearly are patentable without the inclusion of other external hardware. For example, an encryption program that receives a software key and reversibly encodes data in a way that allows the data to be decrypted using a separate software key is certainly patentable. Audio and video compression algorithms like MP3 have long been patentable, as have data processing techniques for removing noise from scratchy signals. Auto tune processing techniques that let singers of living room karaoke sound like Andrea Bocelli — well, maybe almost like Bocelli — are patentable.
The difficult cases arise where the software produces an intermediate result that is not in itself completely useful. For example: As a silly but true example of my personal OCD, I have an algorithm running on an over-automated control system in my house that checks my garage door every night at 7pm and shuts the door if it is open. (I have a bad habit of leaving the garage door open when I come home from work.) But because one evening I was leaving for dinner exactly at 7pm, and the door shut on the top of my car as I was backing out of the garage, I programmed an algorithm to test the systems of my house to see whether the door was safe to shut.
For example, the system looks for movement in the house, the recent status and switching of lights, the temperature of the rooms, the movement of images on the garage cameras, and about a dozen other things. After running through this long list of system checks, if all is good according to the algorithm and no one seems to be headed for the garage or moving a car, the garage door will shut.
In this simple example, it seems to me that the go/no-go decision that the door is ok to shut without killing someone is not that useful by itself — it’s an intermediate result. To actually get a patent on this garage door safety algorithm (a true waste of a patent filing fee) I would most certainly need to include a step in the algorithm where the door actually gets shut or not. I believe the inclusion of hardware – that is the garage door, would be a critical element of this software patent.
How do you advise navigating the fact that our system is now ‘first inventor to file’ and not ‘first to invent’?
My advice is to file early and file often. Provisional patent applications can be an efficient way to protect an invention for a year, while the commercial details are being worked out. If you write the provisional patent application yourself, the cost to file is only $70-$140.
Also, with a provisional patent application on file you can advertise to the world that you are patent pending.
America’s wide lead in venture capital funding is fading and shifting to Asia. Does this have to do with our weakened patent system?
The real story here is that the global availability of venture funding is increasing. Venture funding in the U.S. has been rising over the last half-dozen years at a healthy rate of a couple of percent a year. It’s been steady and continued growth.
To the extent that America’s lead may be diminishing, it’s not because America has been investing less, but because of the dramatic growth of investment in Asian markets. This is all good news for startup companies, including software companies looking for investment.
Asian investment can come in many forms. There are an increasing number of individuals in the U.S. and Canada that have Asian family wealth they want to invest. There are large technology infrastructure companies looking to invest in new products and capabilities, and it’s becoming increasingly common for Asian manufacturing companies to partner with smaller companies to provide investment in the form of manufacturing services.
One last thought about your reference to our patent system being weakened: For the 30 or so years that I have been practicing patent law, I’ve been reading headlines that complain our patent system has been broken or is dead.
The patent system, although more than 200 years old, continues to breathe and live and change roughly in unison with our society’s creative needs. Although the system may seem a little out of sync at times, I assure you: It is far from broken.
Patents continue, as always, to level the playing field for inventors and make it possible for individuals tinkering in their garages to protect great inventions and maybe build great companies.
Patents still make bigger futures possible.
Originally published on Forbes.com June 28th 2018.
