Hike in the fees of IIT and preposterous reasons for supporting it

So Mr. Ramgopal Rao put up a long post in Facebook mentioning his reasons for supporting it. It can be found at: https://www.facebook.com/rao.ramgopal/posts/10218188716681820?hc_location=ufi
Lets go through it one by one.

1. Many top universities in the world have strong Bachelors programmes and strong Ph.D. programmes. The universities which are running large masters programmes abroad are looking at these as revenue earners and charge a huge fee. The streams have also been modified and made contemporary like Cyber security, AI etc. In all top universities abroad, while doing Ph.D., many students get a masters degree along the way. This is how masters programmes have changed all over the world in the last decade.

Yes they do. Some universities in California do it. TU Delft and UTwente in Netherlands do it. But it does not mean its a nice thing. However there is a difference. Netherlands does not charge the high tuition fees from its own citizens or citizens of other countries in EEA (it makes sense as well; the people of those countries pay tax into their own system and its fair that they get a share of it). As for American universities, most do not treat its students as cash cows.

2. Students who come to the M.Tech programmes in IITs are not coming to become academicians. Those who wish to become one, can directly register for a Ph.D. All IITs allow a direct admission into Ph.D (with a high stipend and almost zero tuition fee) without M.Tech as a requirement, including the PMRF. Eventually, Ph.D. students registered in IITs will be able to receive a masters degree once they complete certain credits. Even if it is not so today in some institutes, I am sure, it will indeed become a norm eventually to make the Ph.D. admissions attractive.

Being interested in a PhD does not work like that. I was interested in PhD before I went for MS and then I realized that I am not. I liked working in groups but in PhD, you are expected to do things solely on your own. Also PhDs does not always mean a career in academics. A lot of highly qualified PhDs are required in the industry as well. Look up two new startups in Belgium: Pharrowtech and TuskIC. Both of them are founded by people with PhD degrees.
The way it works is you start doing your Masters, and then during your thesis, you will know if you want to continue in research or go to the industry. By increasing the tuition fees and forcing students who cannot afford to go the PhD way, PhD programs will see more drop out.

3. Students who join the M. Tech programmes in IITs are joining these programmes to find a job, which is fine. IITs therefore need to align their M.Tech programmes to the market requirements. Because of government freely giving fellowships, many institutions are running outdated masters programmes in streams which are not even seen as current by the Industry. In the current scenario, there is no strong motivation for any one to change. Many students take admission into the M.Tech programmes and over 50% of them quit as soon as they find a job. Among the remaining, many prepare for other competitive examinations and have no interest in what they are pursuing. Country’s precious Resources (IIT seats and government funds) are wasted in such a scenario with no benefit to any one. Today, we have uninterested students studying in outdated programmes which are completely disconnected with the market demands. Over 50% dropouts in masters programmes today in IITs only means that, students value the jobs they get after their B.Tech more than the career they can build with their M.Techs. How can we spend tax payers money to offer free education to them, when they themselves see no value in such an education? This is an alarming situation we have gotten ourselves into and obviously, a bit of a reset was needed. I am happy that Ministry and Council took such a bold decision.

If outdated coursework is the problem, then how does increasing fees help? If students don’t understand the value of doing a Masters program, its a different problem altogether. It means that the country does not give a higher preference to people with a Masters degree. And that is symbolic of a lack of highly technical R&D jobs that require an advanced technical degree.

4. If students are willing to pay 20 lakhs for an MBA degree in an IIM, they can surely pay 4 lakhs for a masters degree in IITs. Nobody drops out of IIMs since they see a value in IIM-MBA programmes. On the other hand, we see over 50% dropouts in IIT- M.Tech programmes. After the Council decision, since IITs now need to attract high fee paying students, I am sure, institutions will also align themselves to market requirements and will have a serious relook at the masters streams they are offering. This will bring in accountability at all levels. In the long run, the masters programmes will become contemporary and will train students for the jobs which exist in the Industry.

Business courses are so expensive because they require not just highly educated faculty members but also people who have a good clout in the industry. And to draw the latter talent, you need to pay them high salaries. Are they willing to pay the professors in IITs those high salaries?

And what is this constant talk about market requirements? The main focus of a Masters program is to teach students the basics and how to use them in a real life use-cases and some exposure to industry grade tools. The latter part keeps changing.

Masters program are not meant to teach students how to use the Eclipse IDE but how to take advantage of multi-threading.

5. IITs will focus on B.Tech programmes and Ph.D. programmes like what all research intensive universities worldwide do. Masters programmes in IITs will train students for jobs and those who are interested in academics, will switch over to a Ph.D. The top GATE scorers will still be offered admissions into M.Tech programmes directly as Teaching Assistants and will be offered fee waivers and stipend. However, the fee waivers and stipend will be linked to their role as Teaching Assistants. Others will get into these programmes paying a fee by taking loans, if needed. Since the programmes will get aligned with market requirements, they will be guaranteed a job and can repay loans.

Good researcher ! = Good teaching assistant.
And I studied in KU Leuven. It has been consistently ranked in the top 50 universities. It has been ranked most innovative university in Europe for three years in a row. Its also 600 years old.
And the attitude there is never that Masters program is for jobs. I am pretty sure its not the case anywhere else in the world.

6. The Council resolution also says that all needy students will be given fee concessions and scholarships similar to the B.Tech students. No eligible student in the B.Tech programmes has ever been denied admission in IITs because of her/his poor financial background. There is no reason to believe that this will be any different for top GATE scorers seeking admission into the M.Tech programmes. This is our moral and social responsibility

No, it just makes life difficult for students. It also puts them at the mercy of the administration. Education should be a right and not a charity.

7. The freebies that we have been doling out with tax payers money, to students who are uninterested and into systems which have become unaccountable, needed to stop at some point. Let’s not look at this from an individual’s perspective but rather see it in the larger nation building perspective. IITs are now taking loans from HEFA for their infrastructure. IIT Delhi needs to repay some thing like Rs. 580 Crores to Canara Bank over the next 10 years. Same is the case with IIT Bombay and others. There is a great need to increase our internal revenues.

Which means that these universities are not able to draw funded projects from the industry, which means that the ecosystem of doing state of the art does not exist in India or the quality is not good enough to command the top dollars that foreign varsities charge, which means that the faculty is not good enough. But of course, lets put the burden on the students.

8. IITs need to produce a large number of high quality Ph.Ds who can go out and build top class institutions, get into entrepreneurship and build research culture in our industries. We need to attract top GATE scores into our Ph.D. programmes through full fellowships and scholarships. The limited funds we have need to be diverted into these Ph.D. programmes. M.Tech programmes for a large number of GATE qualifiers need to be self funded (with a guaranteed job at the end of the programme) or sponsored programmes by industries. IIT Delhi has been successfully running for some time one such M. Tech sponsored programme in VLSI. This VDTT programme is entirely funded by semiconductor industries. This M.Tech programme still attracts top GATE scores but with the fellowship paid by the industries.

Yes we need more of such programs but it does not mean that we make life difficult for all Masters students.

9. Industries need to take ownership of M.Tech programmes in IITs and not leave it to the government. It is in their direct interest. M.Techs have indeed played a significant role in building the nation by serving our public sector units and private industries such as TI, Samsung, IBM, Intel, GE, Micron and many many others. Time they start investing their funds in building these M.Tech programmes in IITs and also work with these institutions closely to make the M.Tech programmes relevant. The modified CSR rules also allow this now. We are more than eager to work with them. We can then use our limited resources for doing other things.

Educating people is in the interest of the country. A highly educated population means crime rates are lower, people tend to take better care of themselves etc.

And the reason these companies don’t invest in India is because they don’t yet see the value. Companies always do things that suits them. If investing in India means they will get the best, they would do it. If not, then they won’t.

Using thermal cameras for detecting water leaks

I came across a nice article where they talked about using thermal cameras for detecting water leaks. It instantly reminded me of the project I did as part of a course called Multimedia Technology and Coding. I was given to write an paper on thermal camera and how to use it to detect faults in thermal insulation. There was another part as well where we had to decide on what standards to choose to store the images.

Anyway, this article published in Electronics for you talks about using thermal cameras to detect faults in a swimming pool plumbing system. Typically, they would have needed to dig the entire pavement around to find which pipe was leaking. But instead they used a thermal camera find faults in the water plumbing. The sun heated up the pavements but water having a higher specific heat capacity caused some parts of it to remain slightly cooler. Using the contrast they could find the faults.

Nice use of thermal cameras.

One of the many signs of a slowing economy

So I came across this article in Times of India today. https://m.timesofindia.com/business/india-business/edu-loans-in-india-shrink-25-in-4-yrs/articleshow/69490291.cms

The article is basically a summary of the non-performing assets in the education loan sector of India.

As per the article, the highest portion of NPA is in the sub-4 lakh loan. These are the loans that can be given without any collateral. There can be two reasons why the default rate is the highest in this sector

  1. Because there is no collateral it can mean that the punishment for not repaying is not there and students don’t take it seriously enough.
  2. There are not enough jobs.

I think the last one is more likely. I have both ancedotal evidence and some statistics.

The ancedotal evidence comes from my friends who are trying to set up factories in India. Whenever they start the construction, they get a lot of requests from local people for jobs (one of the condition for being allowed to set up a factory is that they employ the local people). But it’s not always possible. They then demand that they get all their raw materials from the local “syndicate” which is just another name for a shop supplying raw materials for construction. These are typically run by local youths who are not employed.

And the statistics comes from the rumors of a report that unemployment is at a 45 year high. Of course, the government never let the report come out to the public (but will they publish the report now?).

The students who take the small loans are typically for courses like B.Ed, nursing or the many courses in technical training institutes. There is a lot of automation happening at the factory level.. Some people can argue that AI is going to take over these jobs, but honestly that wave still hasn’t come and the data is from the last 4 years.

The default rate in the higher loan amount sector is much lower. These are the ones for things like MBA and engineering courses in one of the better institutes. MBA and engineering graduates can still find jobs although the quality of the jobs and the life satisfaction from it is dropping fast.

Not everyone gets into Google or Amazon or Goldman Sachs or Morgan Stanley OR Oracle or Samsung.. There is a sea of smaller IT companies which are trying to compete with the big fishes. The big fishes are going to become bigger because they have the capital resources to invest into things like AI, which helps them grow much more rapidly than the smaller companies. It’s a deadly positive feedback.

In short, the data presented by this article paints a bleak picture of the condition of the youth of our country. And it’s the beginning of a long list of problems that the new government will have to address. Else, our country will be submerged into chaos.

Another week started, and about my internship at Nokia Bell Labs

And classes resumed, without much of fanfare. Nothing new there, except for the thesis. I have never been so lost in my thesis. Also I had to finish my report about my internship at Nokia Bell Labs. It kind of sucks because we are graded on our internship which means we have to write a report and make a presentation of the work that we did.

A typical internship is 6-7 weeks. 8 hours per day. So anything between 240-280 hours spent. Writing the report is another 10-15 hours. Lets take 10 hours. So roughly 290 hours. Another 5 hours for the presentation. So in all, we have to spend 300 hours on the internship and all associated activities. One credit roughly equals 25-30 hours of work. So 300 hours should be anything between 10 and 12 credits. But we are only given 6 credits. [Source: https://ec.europa.eu/education/sites/education/files/ects-users-guide_en.pdf ]

For me this is really annoying.

But lets talk about the internship. The people in Bell Labs are nice people. But the work I did there was more like research as opposed to actual analog design. Honestly, that is fine by me. I worked for three years in a major analog design firm and I have had enough of the “experience” in a huge multi-national company. But what I did not like too much about the internship was that the specifications were not well defined as the beginning. Which makes it really difficult to work on something. And I was sick as well, but I had to push through. It is something I did not let the people there see.

But I think the star of this blog was Brecht Francois. A brilliant guy, with a PhD in power amplifiers from KU Leuven (his guide was Patrick Reynaert who is the top authority on power amplifiers, especially in the mmWave domain). He understood the limitations, had good discussions and it was really nice working with him.

So will I ever go back to Bell Labs? Well yes, if I can get to work with Brecht.

Unfortunately, I cannot talk about my work there (for confidentiality reasons). A big shout out to Jie Yang who was also interning at Bell Labs and was from KUL. The only friend I had there really.

One year at Leuven

Note: I started this article way before 15th of September and took several days to write it.

Although I officially complete an year in Leuven on 15th September, I thought I might pen down a few lines on how my year has been now. I have been told that there are two ways to write a review: compare your experience with the expectation that you had set out with; or write your review completely disregarding the idea that you had before you embarked on the experience. I, however, believe that a bit of both is necessary. This stems from the fact that I am too young and too less traveled to form an accurate idea of living in a foreign country or study in a top ranked university.

Now to give some background about myself, and why I left a very high paying job at Cadence Design Systems or a job offer at Samsung Semiconductors to pursue a Masters. Back in 2013, when I got my job at Cadence Design Systems, I did not study much for GATE and rather started working on various university events. But, thanks partially due to luck and partially due to my excellent concepts in the subjects, I secured an all India rank of 46 in GATE 2014. Those of you who are familiar with it, would know that even a rank of 46 is not enough to guarantee a seat in the MTech programs (Microelectronics) either in IIT Bombay of IISc Bangalore. I had no interest to pursue a MS program in India. Plus, the team in which I got into at Cadence had some of the best analog design engineers in India.

Anyway, during my job, my colleague and a good friend of mine, Gunjan Mondal asked me to get a MS. I also talked to other colleague including Vishnu who told me, “If not anything, you will enjoy the two years. These jobs will always be there for you.” Sunil Rajan, another brilliant mentor of mine recommended the same, “See in a Masters you can work on one thing exclusively for one year; you don’t have to do half the crap that you do in the industry. You will learn a lot.” And finally my boss, SKG told me, “You are young. You should always chose a path where you learn the most.”

At the time I was confused between KUL (offered me a full tuition fee waiver), TU Delft (offered a partial tuition fee waiver and a part of the excellence scholarship) and UTwente (full excellence scholarship). They all told me, if money is not the constraint, go to KUL. And here I was.

I did look at the subjects that I will have to study at KUL. It had a lot of circuits (which was fine by me) but also antennas and EM propagation. I never took a formal course on electromagnetics in my Bachelors and it was kind of scary. (But I still read some senior UG level textbook while I was working so I knew it won’t be mighty difficult).

I remember the first class. It was by Patrick Reynaert. Design of Electronic Circuits. The course covered operational amplifier design. There was a project (to design and layout an OTA; I was kind of lucky that I was allocated specifications that was not possible to meet. When you have to prove that something does not work, you will have to put in twice the effort to convince someone and you learn twice as much). Prof. Patrick is an exceptional professor who really taught the students a systematic way to design most of the common single stage OTAs.

And then we had Design of Digital Platforms. The project was really time consuming and kind of difficult. I was lucky to have a partner like Luis Diego. He worked several years in the industry, and was good with Verilog and we managed to finish the project and get a decently high score. We had a completely useless partner in <<name deleted>> who never worked, lied and tried to manipulate. I believe that none of my batchmates who teamed up with him in different projects had a good experience. (Lesson:Always select good project partners!)

The most difficult course was Digital Signal Processing. The course material was hard, and the project was not hard but time consuming. But then when you saw the finished project, you cannot help but remark that it is indeed a good learning. We had to transmit an image (of the professor; but I also did the same with my girlfriend’s picture) using audio. The modulation used was OFDM.

There were other courses like Analog Building Blocks for Signals and Systems. It was a great course for two reasons:

  • One, you learned of the top level specifications of analog blocks, what it means and some of the skeleton architectures of the most common analog blocks.
  • The exercise sessions were really good.

There was also a course on transmission lines and design of microwave amplifiers. In the final exam, we had to design the matching network for a potentially unstable amplifier. Its taught by Prof. Bart Nauewelers who is really nice professor.

Exams in KUL are held in January (first semester), June (second semester) and August (re-takes of courses that you may have failed in the first or the second semester). The exam periods are typically hectic. Other than the usual exam time late night stories, the only memorable part of it was my bed broke. There is this supporting frame on which the mattresses sit, and one of the ribs snapped. I knew I could fix it through some duct tape and wooden beams. So I went to my landlord.

“Yes Vivek, how can I help you?”

“Umm, my bed broke.  I was wondering if you had some duct tape and a wooden beam to spare.”

“Atleast tell me the girl was blonde.”

Despite what a lot of people thought, it did not break due to any girl. But anyway it was fixed and the rest of exam period went without any misfortune.

The second semester was rather dull as I had to take courses which were not that interesting . This was partly because I had to take Dutch classes, which was held twice a week from 7pm-9pm. It meant that two evenings in a week were completely wiped off.  I liked computer architecture.(The TAs were really great). I liked Design of Analog Integrated Circuits. I sort of liked antennas. I hated EM propagation. Technology fr Microelectronics is a course that gets special mention.

I also became homesick during my second semester.

Although I thought the most interesting part would be the project where I had to design a DCF77 receiver, it turned out to be rather boring. I was responsible for the digital part and it was okayish experience. Towards the end, I looked forward to the sessions because there was one TA I really liked in particular and we got friendly.

A better experience was with the project on Design of Analog Integrated ICs. We had to design a variable gain switched capacitor amplifier. The TAs were nice and because they were all international PhD students with whom we could connect more, share our frustration of the lives in Belgium. See the thing about KUL is that the classes are really small (in number) and you need to kind of bond with the TAs to have a good experience. Having empathetic TAs help. International people don’t have a home they can go back to on weekends. They need these little anchors to help them survive KUL’s MSEE program. It is nowhere close to being easy. I became good friends with one of them (Dileep).

Anyway after my second semester exams, I had an internship at Bell Labs. Unfortunately, I cannot talk of the work I did there but it was really nice working with Brecht Francois. Brilliant guy. Overall, Bell Labs was a nice place to work.

I returned to Belgium last week, and I joined IMEC for my thesis. As of now I am enjoying it and I hope to do good work on thesis. Lets see how my second year goes.

Blok period, and all the contemplation that comes with it

One of the most interesting phases of student life at KU Leuven is the blok period. It is the two weeks before the final exam (we only have one exam in a semester anyway) where you try and go through the subjects that you did not go read the entire semester because you just hated them!

But anyway, blok period for an electrical engineering student is really crucial. Most of our time during the semester is spent on projects. (Electrical engineering projects are extremely demanding, but you go away with so much of knowledge!). And this really the only time we are left to study properly for the exams.

I ran into Sujoy, a post-doc scholar in the COSIC group of ESAT yesterday at the Carrefour market. He was the TA to a very interesting course last semester, ‘Design of digital platforms’. We were just normally talking when he advised me to diversify my courses (read: take more fundamental maths and physics courses) rather than take more and more circuit design courses. Any new research field or any new novel idea is really an implementation of the basics.

And today I was reading Antenna. A dreaded course. Dreaded because my basics in electromagnetics are not that strong because I never had a good teacher in my bachelors university. Then a thought came to me: I understand (for most part) the material that is covered in the slides. But, today if you ask me to design a point to point antenna, will I be able to do it? Honestly, no. Given a current distribution, I can use Green’s function to calculate the magnetic vector potential, and then use it to derive the magnetic and electric fields. And then derive the impedance. But that’s all in theory. I don’t think I am confident enough to be able to build an antenna. But then, I have been taught the basics of the mathematical tools to do antenna design.

And it hit me. If today, I am extremely good in electromagnetics, I can pick up antennas much more easily than the other way round. If I look at analogue circuit design, if understand the basics of electrical circuits and feedback really well, I can go ahead and design a circuit. I don’t need to know a hell lot about MOS or a BJT to begin with a synthesis problem. If I am good in mathematics, I can be good in algorithms. Case in point: I studied a lot of mathematics while I (used to) solve problems from project Euler. I came up with algorithms and the formal proofs of them from the first principles.

In Belgium, they focus a lot on the basics in the first years of their engineering degree. Those exams are hard. Only after the basics, they introduce them to ‘engineering topics’. This was never the case in India. The basic courses were too easy, and the teachers who taught those basic courses were from the Faculty of Science, and they could not give a shit about the engineering students even if they wanted. That really stunts the ability of future engineers to be able to solve synthesis problems. If you are smart, then analysis problems can always be solved. But to solve a synthesis problem, the basic mathematical and physics tools need to be sharp. This is clearly lacking in the Indian education system.

Ok enough rants! Let me get back to my books! *Fuck!*

India’s increasing expenditure in defence

Ever since Narendra Modi came into power, he has been trying to push for promoting manufacturing in India. India’s GDP for too long has been dominated by the service sector. Here is a snapshot of a table [1].

gdp_sector_wiseA big contribution of our GDP comes from the service sector, and it has been growing over the last few decades. I think over-reliance of the GDP on service sector is not a good sign as the decision making powers lies completely with the organizations that ask for the services in the first place.

Narendra Modi in his bid to reverse this started the Make in India program which focused upon building manufacturing capabilities inside India, but it never really took off. He has now gone back to a more popular method to infuse enthusiasm in the manufacturing sector: defence.

The reason why engineering and science progressed so fast in the beginning of the 20th century were the two world wars. Communication and radar technology development accelerated during those days. The top institutes in US at that time played a major role in radar and communication systems for the aid of military. And it always easy to get funds from the Congress by saying , “We need more funds because we are researching this area which can help us in better equipments for the army.” Cold war helped too.

India is the world’s largest arms importer. Which is also not a great thing if a country aims to become a super-power. India needs to modernize its military inventory.

Combining the two things, the current government is tying up defense purchase contracts with building of a local ecosystem. By forcing the major arms manufacturer to build their assembly lines here in India, and to use 30% of the materials sourced locally, they have no option but to build a local supply chain to bag the >100 billion dollar contracts. Which means we will see a lot of the suppliers who have traditionally supplied to the big sharks moving (parts of) their manufacturing shops here.

When an ecosystem starts building up, we will definitely see a lot of local businesses trying to compete to get a piece of the pie from the likes of Boeing and Lockheed. Indians are fiercely competitive, and have given several deep pocketed companies like Amazon and Uber sleepless nights. I see no reason why it would not be the case here provided they get the necessary funds! [Which has become a bit of a thorny issue with all hundreds of crores of write-offs the banks are having to do thanks to non-performing assets].

I believe its a good start for the Indian economy as a whole.

On a more related news, the government has imposed 20% import duty on printed circuit boards assembled outside the country. This means that on an average phones are going to get expensive. Indians love free and cheap stuff. This is a nice trick (without becoming a trade war) to force the companies to assemble the PCBs here in India. I am kind of happy about this move!