Patent issues in Indian Engineering Education

Counting patents, done the wrong way, misses real innovation

Driven by industry demand, engineering colleges have mushroomed in the country over the past two decades. While their number has grown multi-fold, the quality of education delivered has gone down. Government has tried to improve this situation in various ways, by making accreditation metrics stronger for colleges and universities. Number of patents is one such metric. This may look like a great way to measure innovation. However, an initial analysis shows that this has become a perverse incentive, resulting in actual harm.

NAAC(National Assessment and Accreditation Council) is an autonomous body funded by the UGC(University Grants Comission). NAAC assesses and accredits Higher Education Institutions. Engineering colleges are providers of education and skilling services. In that sense, they serve two customers - students looking for high quality education, and private sector looking to recruit skilled engineers to drive their business. For private engineering colleges and universities, NAAC accreditation is important, as it a recognised indicator to assess the infrastructure, instruction, research, and learning standards of a university/institution. NAAC accreditation also provides access to funding programs from Union and State bodies.

NAAC accreditation looks at the number of patents published/granted both. Here’s a snippet from a relevant document¹.

To gain/maintain accreditation, patents are a useful tool. Colleges are likely to be incentivizing patent activities by faculty and students to maintain and improve their rating. This may look harmless at first glance, but we'll soon see why that is not necessarily the case.

Filing patents in India costs Rs 1600 for individuals, startups, MSMEs and Educational Institutions. Getting a patent grant is an elaborate process, taking significant time and money. However, "publishing" a patent does not incur significant costs, and can be done in a very short time frame. Publishing thus generates valuable metrics at a low cost, in a timely manner.

Analyzing patents filed by academia on the inPASS(Indian Patents Advanced Search System) reveals various disturbing trends². A large number of patents have been filed around the areas of AI and IOT, about two thousand filings this year alone. Most of these lack the requirement of a novel inventive step that shows industrial application, clear attempts at rehashing engineering projects as patents. Most of these actually fall under the bucket of “software patents” as well³, and will likely never be granted in the end.

Globally, the number of inventors per patent hovers around 3. In these patents, it is common to see anywhere between 4 to 12, to even 15 inventors - comprising mostly faculty and students. Whether this is being done to share costs, or credits, or both is something that needs a deeper analysis. There might be some motivation to share costs if the overall costs involved in a patent generation are higher. There have been anecdotal cases of agents providing services of creating and filing patents, but it is not yet known how widespread this is. Usage of services of agents would drive costs higher.

Some of the patents have seemingly unlikely collaborations. Wide collaboration in a single patent, say spanning Karnataka, Tamil Nadu, Odisha and Telangana, must raise red flags. NAAC has requirements around collaborations to gain/maintain accreditation. Real collaborations take time, effort and cost to build. Buying the services of agents who sell collaborations is likely to be a surefire way of achieving the metrics at a low cost, to the detriment of the system.

In the business world, patents are filed as part of an elaborate strategy. The publish step is generally delayed as long as possible. This is important, as publishing a patent gives a heads up to competitors about what's coming next. Compared to this, most of the academic patents seem to be "published" almost as soon as possible. Thus, it becomes harder to argue that there is an intent of commercialisation of the patent in the academia - which is the part of the reason why patents exist in the first place. In 2018, it was reported that VTU (Visvesvaraya Technical University) - Karnataka's university which has maximum engineering colleges - earns nothing from patents. For perspective, all the IITs combined earned less than 1 million US dollars from patent licensing⁴. Even at what we consider exceptional levels, the outcomes have been poor. Cut to 2023, the numbers might be better, but the scenario isn't likely to be any different.

The large number of such filings is likely to have taken a silent human toll already. Educators in the system would have reasonable knowledge of the systematic gaming. But they would be helpless, as any evidence they can produce is likely to be deemed anecdotal. In a system that takes a long time to resolve complaints, complaining against a colleague isn't exactly a wise strategy either. The number of students involved in these patent filings is also a cause of concern, as their world-view is being shaped by their experiences.

Patents "published" are overall a poor metric to track, and are causing immense harm. NAAC rules must get rid of the patent "publish" metric, and keep the patent "grant" metric. This is more likely to drive useful long term impacts, and reduce gaming by players in the system.

1. NAAC - Standard Operating Procedure for Data Validation and Verification (University)

2. Note that this analysis is limited to IT/software patents, an area closer to the author's expertise. These form a large chunk of seats and courses in our higher education system.

3. According to Indian Patent Law, section 3(k), computer programs “per-se” and algorithms are not patentable

4. Indian Express, IIT-Bombay earns highest for patents