The Lost Code

Actors of software, particularly from colleges, are malnourished artistes.
Lack of adequate guidance and orientation often spoils the play; and when things
look a little bright, they fall short of being beautiful because of little
protection against villains. The software a student writes and burns on to a CD,
is attached to his project report and, in some cases, filed in a library. It may
or may not be available to newer batches of students. Little is done to document
and host it; lesser still are the efforts to identify a market potential, bundle
it into a product or sell it. Essentially lost is a revenue stream for the
college and a feather in the student's cap.

Filing for patents and looking to earn from patent licensing is something
most engineering institutions in our country have opened up to only recently.
But no study has been conducted yet to judge the market value of the software
which students write. What students basically develop are tools-that help in
design and analysis, among others. It might not be appropriate in terms of
looking at them as products. But these utilities can be packaged much better and
made available. Let's say someone from an IIT creates a simulator, which plays
a specific role of simulating a wireless LAN. That software can at least be
packaged in a reasonable form and made available to some of the universities-which
itself should serve a huge community. "One of the things I am concerned
about is that even this is not happening. Even within the student community,
these kinds of software are not changing hands. When it happens, people will
become aware and start plugging in their IP claims," says Gopi Garge of the
department of Electrical Communication Engineering, IISc.

Pinning it down

What is therefore required is a certain level of awareness. The premise
should be to document and host whatever is created, good or bad, so that people
take notice. Then, there is always a possibility, an opportunity to strike big.
"You will perhaps not come up with a billion-dollar product but, who knows,
maybe a $500 product. A lot of people will ask questions about how much time and
energy will be spent in packaging this versus its market value. But it is not
just a question of packaging it," says Garge.

Many
feel there is a lot of good software languishing. But one professor who is not
too sure about quality is NJ Rao, chairman of the Department of Management
Studies at IISc. Many a time colleges have cried out their readiness with
products, only to falter when the industry first tests them, he says. "The
proper structure is not incorporated at the design level. Neither the faculty
nor the students have the time to do it. If the practice of implementing a
project is done in a right way, then the amount of output can phenomenally
increase," he says. Why this happens is Rao's guess: while knowledge
pools exist at various places, it is not internalized in any college.

To the optimistic lot belongs Dr Mary Mathew of the same department. There is
good commercial potential to market students' software, but a lack of
awareness about what could be called their own IP, she feels. There is also a
second, and perhaps, a more serious problem-Indian institutes cannot negotiate
IP ownership with their clients-companies that sponsor R&D projects in
colleges want complete transfer of IP ownership as opposed to a situation where
the copyright is shared. IISc is beginning to do that now, and from here on,
things should look brighter.

But for a start, Mary suggests that colleges set up knowledge repositories,
wherein software deposited is available for reuse. The learning comes from
corporates who already practice that for inhouse software development. Second,
would be to identify good software that can be patented. The code will probably
be written in a tangible form. The degree to which it has to be strategically
patented further has to be verified on certain parameters-whether proprietary
algorithms are being developed, or if R&D is involved in its development, as
well as the design architecture required for a certain OS. "If there is
something potentially patentable, the professor would know. To that extent,
there should be a healthy relationship between the teacher and the student or
this can be missed," she says. But ultimately, it is the institution's
responsibility to file for the patent.

The need for an IP culture

IISc has an IP Cell that professors and students approach. All engineering
colleges, Mary feels, should replicate this model and promote IP practices among
students. For example, an important aspect of IP practices is prior art
analysis. Students have to be taught this or they usually end up replicating
software which someone else developed. "If they have to get to the level of
leadership and work on areas that are frontier, it is good for them to go
through areas of prior art analysis-and know what has been developed before. A
lot of IP culture is now coming into our university system," she says.

A college should ideally have an IP Cell and a Technology Licensing Office (OTL).
The second one is quite popular in the US and is set up for the
commercialization of IP relating to student tech endeavors. It functions as a
business unit, accessing the cost of the patent, negotiating, looking at
diffusing it into the market. IP cell would primarily look at promoting filing
of patents and IP culture. Bringing about cultural changes would include
changing the way software projects are designed. The final specification of the
software has to be defined at the beginning. "If you can convince the
student that this is the most important step in his professional life, the
industry will gain fantastically," he adds.

The industry, we know, is not blind. It will pick up good work from anywhere
and if that happens to be a college, young actors will have great roles to play
and the producer, more money.

Goutam Das in
Bangalore

Founder and CTO of Beceem Communications, Arogyaswami Paulraj

Founder and CTO of Beceem Communications, Arogyaswami Paulraj, is also a
professor of Electrical Engineering at Stanford University, working in the area
of mobile communications. He tells Goutam Das of Dataquest that India has a long
way to go as far as protecting IP is concerned.

What happens to the IP generated at engineering institutions in India?

Some PhD level projects may generate IP (ideas, algorithms), but top
universities do harvest it. Tradition of getting patents and licensing IP is
still low-very few, say about 20 universities (IIT Chennai in Wireless is an
example), are at a stage where their IP is relevant and is eagerly sought by the
industry.

Is there a need to package and market the software written by students?
What happens to the software written at Stanford?

Commercial products and academic research do not mix well. No production
software is written at Stanford. Sometimes, we generate some software tools such
as SPICE and MATLAB, which are usually not production quality and it is made
available in public domain. Industry may license and market these as products
after adding value and support.

What are Stanford's revenues through patent licensing?

Stanford probably files around 200 patents per year. Our annual revenue from
patent licensing is about $200 mn a year.

How can the IP generated be protected and better projected?

Stanford has a large office-OTL-that actively deals with IP protection,
licensing, litigation... If you look at India's share of publications in world
class journals, where the rest of the developed world publishes-say for
example IEEE, our share is very low. I think well below 1%. Normally, IP value
and publications record in such journals are highly correlated. So we have a
long way to go.