Patent protection is best suited for inventions that are at risk of being publicly detected or independently developed. Innovation that will likely not become publicly available or would require significant work to recreate may be a good candidate for trade secret protection.
Today’s Robotics Innovation Landscape and the Role of IP in The Field of Robotics
Nicholas Terrell and Shannon Hughes Mastick | MARSHALL GERSTEIN
Innovation in the field of robotics can take many forms – from mechanical components to electrical controls to software, and even processes of producing components. Although intellectual property (IP) law applies to all of these forms, some categories of IP protection are better suited for certain types of innovation. Generally, patent protection is best suited for inventions that are at risk of being publicly detected (either directly or through reverse engineering efforts) or independently developed. Innovation that will likely not become publicly available or that would require significant work to recreate may be a good candidate for trade secret protection.
Patent Prosecution and Protection of Robotics Innovations
Because it is difficult to hide physical components or software shipped to a customer, many aspects of innovations in robotics may benefit from patent protection. Generally speaking, patents can be used to stop competitors from making, using, selling, offering or importing a patented invention. Consequently, by patenting your robotics innovations, you may be able to establish a competitive advantage.
Utility patents may be used to protect new, useful and non-obvious inventions, as well as improvements upon existing inventions. For instance, it is possible to patent innovative mechanical features of robotics innovations as well as innovative techniques for building or manufacturing robotics equipment. Moreover, it is possible to patent innovations in techniques used in robotics to perform various tasks, as well as unique systems of robotics equipment and how various elements of the system interact with one another. Additionally, it is possible to patent steps taken by a computer algorithm that improve the functioning of a computer or device.
Unlike trade secrets, obtaining a patent involves a formal application and review process (referred to as “patent prosecution”) prior to being granted a patent. In the United States, the U.S. Patent and Trademark Office (USPTO) receives, examines, typically rejects and eventually allows a patent application before issuing an enforceable patent. Patents are only enforceable in the jurisdiction in which they are granted, so obtaining protection in other countries requires filing corresponding patent applications in those jurisdictions. Few applicants seek truly global coverage, instead targeting economically significant jurisdictions.
Prosecution of patent applications for inventions involving mechanical or electrical components typically focuses on defining the novelty of the invention over prior art (preexisting technology) in the field. For software-related inventions, however, patent prosecution will often include further consideration of subject matter eligibility. In the U.S., the subject matter eligibility hurdle for software inventions has been a particular concern since the 2014 decision of the U.S. Supreme Court in Alice v. CLS Bank. Subject matter eligibility analysis at the USPTO typically focuses on whether the invention is directed to an “abstract idea,” while the European Patent Office (EPO) focuses on identifying a “further technical effect” beyond the software per se. Unlike some other areas of AI and software innovation, robotics inventions typically include readily identifiable aspects supporting patent eligibility. In fact, the 2014 Interim Patent Guidance on Patent Subject Matter Eligibility published by the USPTO shortly after the Alice decision gave an example of “a robotic arm assembly” with a control system as being appropriate for a streamlined analysis of patent eligibility. In most cases, software inventions relating to robotics can clear the patent eligibility hurdle by reciting controlling an action in the physical world.
Although less common in the field of robotics, design patents may also be used to protect the visual characteristics or ornamental design of a product. For instance, it is possible to patent the visual characteristics of a user interface display, or the ornamental design of a particular tool or machine.
Trade Secrets Considerations for Robotics Innovations
Trade secrets may also be useful under certain conditions to protect innovation in robotics. Although trade secrets can potentially last indefinitely, they can be lost by the protected information becoming available to the public – even through reverse engineering or by independent invention or discovery of the same information. Thus, trade secret protection is typically better suited to innovation that (i) can be kept away from the public (such as back-end software or manufacturing processes), (ii) would be prohibitively difficult to reverse engineer or recreate, or (iii) is not eligible for patent protection (such as negative information regarding experimental results that proved to be inadequate solutions).
Generally speaking, trade secrets may be used to protect methods, techniques or processes that: (i) derive independent economic value from not being generally known, (ii) are not readily ascertainable, and (iii) are the subject of efforts to maintain secrecy. By holding an innovation as a trade secret through adequate protections and non-disclosure provisions, your business may prevent employees, contractors, suppliers, customers and others from revealing the innovation to the public. Although previously defined only under state law, trade secrets have been protected under federal law in the U.S. since the passage of the Defend Trade Secrets Act in 2016.
Trade secrets can be a useful tool for protecting certain kinds of robotics innovations, especially when used in conjunction with patent protection. For instance, while certain computer software processes and functionality may be patented (as discussed above), the actual source code and certain details of algorithms used therein (e.g., parameter values or equations) may not be patentable, instead such information may be best protected as trade secrets.
While anything that is patented will eventually become public, software code kept as a trade secret will not become public unless it is revealed or otherwise discovered by the public. An additional benefit of trade secrets compared to patents is that trade secrets are able to protect an innovation in perpetuity (until the trade secret becomes public knowledge), while patents only protect an innovation for a term of 20 years (after which the patented subject matter becomes freely available for use by the public).
As an important note of caution, however, trade secret law does not prevent reverse engineering by a competitor. To the contrary, many statutes expressly permit the reverse engineering of trade secrets. In contrast, patent protection will protect an invention from being practiced by a competitor even if the competitor independently develops the patented invention or discovers certain details of the invention via reverse engineering.
The Legal Status of Inventions Derived from AI and Robotics
As artificial intelligence (AI) systems become more and more sophisticated, questions have recently arisen as to whether it is possible for an AI system to be an inventor. Whether or not AI systems are capable of inventing in a philosophical sense, in the patent world the answer to that question seems to be “no” – at least at this point.
According to a recent petition decision from the USPTO, under current U.S. law only “natural persons” may be named as an inventor in a patent application. In the case upon which the petition decision was based, Dr. Stephen Thaler applied for a patent on an invention of “Devices and Methods for Attracting Enhanced Attention” allegedly invented by his AI system “DABUS” (Device for the Autonomous Bootstrapping of Unified Sentience). Although Thaler invented DABUS, he did not invent the specific invention claimed in the current patent application.
In seeking a patent for the DABUS-generated invention, Thaler argued that DABUS was a “creativity machine” programmed as a series of neural networks trained with general information in the field of endeavor to independently create the invention. Thaler further asserted that DABUS was not created to solve any particular problem, and that it was not trained on any special data relevant to the invention. Therefore, DABUS – and not any particular person – recognized the novelty and salience of the invention. However, despite these arguments, the USPTO rejected the patent application because DABUS was not a “natural person” and could thus not properly be an inventor.
The European Patent Office (EPO) and United Kingdom Intellectual Property Office (UKIPO) have also recently rejected patent applications allegedly invented by DABUS on the grounds that only natural persons may be named as inventors in patent applications. Although similar questions of inventorship will undoubtedly be raised in the future, the reasoning behind these decisions appears to be sound. Absent legislative action, it is unlikely that AI systems will qualify as inventors for the purpose of obtaining patents.
About Nicholas K. Terrell
Nicholas K. Terrell, partner at Marshall, Gerstein & Borun LLP, assists clients in obtaining and protecting intellectual property rights through development and evaluation of patent portfolios, with a particular focus on electronic systems and software.
About Shannon Hughes Mastick
Shannon Hughes Mastick, associate at Marshall, Gerstein & Borun LLP, handles all aspects of domestic and international patent prosecution, from helping inventors define and document their innovations to drafting patent applications, responding to USPTO office actions, conducting interviews with patent examiners, and managing foreign agents.
The content & opinions in this article are the author’s and do not necessarily represent the views of RoboticsTomorrow
This post does not have any comments. Be the first to leave a comment below.
Post A Comment
You must be logged in before you can post a comment. Login now.