'It is in electronics that the gap between where we are and where we need to be is most obvious and most persistent.'
'It is not only a national security issue, but also a commercial issue,' argues Rajeev Srinivasan.
Illustration: Dominic Xavier/Rediff.com
The Pulwama massacre and the Balakot strike appeared on the surface to be classic instances of conventional 20th century warfare.
The terrorist suicide bomber was 'wetware': Humans used as unthinking machines.
The Balakot bombing, the aerial dogfight and the downing of both the MiG-21 and the F-16 was hardware.
The psy-ops were Goebbelsian propaganda.
Behind the scenes, however, I could reasonably argue that it was all about electronic warfare.
That Indian intelligence failed to identify, track and intercept communications between the suicide bomber and his handlers is a failure in using pre-crime big data.
Machine-learning systems can, with increasing precision, predict future terrorist attacks by recognising known patterns of behavior of individual terrorists who are being electronically surveilled.
It is also quite possible that communications between Indian soldiers were being done on Chinese-made cellular handsets that are so ubiquitous, and that these were intercepted and sent to Chinese servers.
That is precisely what happened to the African Union in Addis Ababa, whose HQ was built by the Chinese.
For five years, every night, confidential data was quietly beamed from the building to telecom servers in China.
The Balakot strike was a success because the Indian AWACS plane jammed Pakistan's Chinese radar systems, so that they didn't scramble their own fighter jets.
Precision-guided Israeli S-2000 munitions hit the right targets because they had been fed the GPS coordinates of the terrorist training camp, and could maneuver themselves there.
The fire-and-forget AMRAAM missiles fired by the Pakistanis, and the Indian 'kill' of one of those by a Indian air-to-air missile (an extraordinary accomplishment, by the way) were made possible by electronics.
Information warfare, through faked pictures, planted stories, diversionary tactics (the fuss that Masood Azhar was dead) and Internet memes can be powerful.
So can manufactured Internet content intended to create riots or induce panic.
And 'deepfakes' can hit the scene with Indian leaders 'announcing' in realistic video that Delhi and Mumbai are under nuclear attack, triggering mass panic.
Cambridge Analytica has already demonstrated how to micro-target and customise disinformation to you that you will be guaranteed to believe.
Besides, tomorrow's wars may be through massed swarms of drones armed with simple weapons.
Already drones can do damage: I watched a startling video of an Israeli drone, controlled from hundreds of miles away, killing, with a single bullet, a Yemeni militant who was making a speech in an open-air arena.
A 1,000-drone Chinese swarm using AI to mimic a flock of birds (watch a video of a murmuration of starlings) will be impossible to defend against: The equivalent of an irresistible Mongol horde like Genghiz Khan's.
And what will you do if your enemy does not need to fire a single shot to bring you to your knees by turning off your electric grid, collapsing your banking system, opening all your dam shutters, taking over your air traffic control system, or hacking your self-driving cars and your traffic lights to create total mayhem on the streets?
That scenario is here. Today.
Electronic warfare can well be an existential threat.
Countries may be forced to surrender.
What underlies all these scenarios is electronics and semiconductors.
India is woefully behind the curve in these technologies and in related software.
For instance, AI, 5G, and IoT (Internet of things), likely the most important technologies for the near future, are all dependent on three things: semiconductors, algorithms and data.
The US has all three.
China has the last two, but it is vulnerable (remember the ZTE case where a US embargo almost shuttered the company before there was a reprieve?) because US firms have a monopoly on the electronics.
India is in the truly appalling position of being deficient in all of the above.
Our vaunted software capability is in services (development: Implementing a well-defined thing), not in creating unique algorithms (research: Inventing something new).
We have gigantic amounts of data being generated by our citizens, but all of it is being plundered by Facebook, Google, Alibaba and TransUnion, among others.
And we have little competence in semiconductors.
Most of this is because India does virtually no original R&D, in academia, industry or government.
And even if somebody does, and by chance come up with something original, we then contrive to kill it off.
Two examples will suffice: The Reva electric car, and the Simputer.
Both were leading-edge, but the country simply didn't have the wherewithal to make them succeed.
But India does have potential.
We could prevent foreigners from making free with our data, by forcing them to retain the data in India, or to exit.
For instance, there is really no need to allow many dubious Chinese social media apps to exist in India.
Indeed, there is a case to be made for politely asking Facebook, Amazon, Twitter and Google also to leave.
Is it the case that India couldn't make alternatives? Even through rare, there are now Indian apps that are halfway decent: BHIM, for example.
But it is in electronics that the gap between where we are and where we need to be is most obvious and most persistent.
It is not only a national security issue, but also a commercial issue.
It's a fair bet that the large and growing trade deficit with China is mostly because of electronics imports, particularly telecommunications.
The ministry of electronics and information technology (MeitY) estimates that our trade total trade deficit by 2025 will be $400 billion in electronics alone.
Thus, there is a great deal riding on the new National Policy in Electronics 2019 unveiled in February.
Can this revive the moribund electronics industry in India? The outlook is mixed because the policy, even though it is a welcome step, seems to be missing some crucial points.
But then, it is in the nature of policies to be general: It is up to the ministries concerned to put together actionable efforts to make things happen.
The National Policy in Electronics 2019 notes with chagrin that India accounts for a measly 3% of global electronics production, no more than $60 billion.
This has to change.
An earlier presentation from MeitY had identified the gap between production and consumption in a more granular fashion.
It showed that India is doing reasonably well in increasing local production in a few areas: Mobile phones, set-top boxes, monitors, PCs, inverters, and so on.
Alarmingly, however, India is doing much worse in the following between 2015-2016 and 2016-2017, wherein imports have increased substantially: telecom base stations (227%), routers (78%), telecom equipment parts (39%), solar cells and modules (36%), populated printed circuit boards (118%) and bare PCBs (43%).
It stands to reason that India would import more solar cells, as Chinese over-production has caused prices to crash by as much as 80%.
In fact, it is a good thing to take advantage of possible Chinese distress 'dumping' at prices below cost of production to build up India's renewable energy efforts.
Importing of PCBs is also fine, as it means more products are being assembled in India, although we should be producing more bare PCBs, and even populated ones.
It is in telecom that the danger signals are flashing.
In particular, telecom equipment parts imports have zoomed from about $4.5 billion to $6.3 billion in one year alone.
This suggests that India is getting more and more dependent on Chinese telecom vendors, presumably Huawei and ZTE and the likes.
The brouhaha over Canada's arrest of a Huawei executive, and the rejection of Huawei's products by several countries are indications that there is a high level of anxiety about that company.
There are chances that Huawei may dominate the coming 5G telecom race: They are already the largest seller of telecom equipment in the world, and they are accused of being essentially an arm of the Chinese government.
It would be imprudent for India to entrust its next generation telecommunications, which would control billions of Internet-of-Things devices, to a manufacturer that rouses such strong fears.
The question is, why can't India build up its own world-class makers of telecom equipment, as well as influence the setting of standards for 5G and later 6G?
Similarly, why can't India set up world-class semiconductor foundries in silicon and possibly other, more exotic materials like gallium arsenide and germanium? IISc, Bangalore proposed in 2018 a Rs. 2,500 crore project to build an entire gallium nitride fab.
Is this the leapfrog India needs? Let us note that China has decide to invest $47 billion in semiconductor technology.
A big problem with the policy is that it doesn't dare to dream big, perhaps along these lines.
In two seminal articles in the Harvard Business Review in 1989 and 1900, C K Prahalad and Gary Hamel introduced the concepts of Core Competence and Strategic Intent.
A strategic intent is a long-term goal that a firm (or in our case, a country) decides to pursue.
It might even appear a little quixotic, given where the entity is at the time.
But that is not the point: It's not where we are today that matters, but where we can get to.
If we aspire to something really big, it is not impossible to get there if there is a firm commitment.
Prahalad and Hamel cite Japanese firms that in the 1970s and 1980s created a strategic intent to dominate certain industries: examples include Komatsu in earth-moving equipment, Sony in electronics, Honda in automobiles, and Canon in opto-electronics.
Even though their American competitors such as Caterpillar, GM and Xerox were far bigger, by articulating a clear intent (and importantly, encouraging employees to come up with their own initiatives in the pursuit of that goal), the Japanese firms systematically enhanced their positions so that they eventually became global behemoths, and that too in a matter of only 20 or so years.
In the pursuit of their strategic intent, they also carefully enhanced their capabilities in critical areas: These are the Core Competences that they can execute better than anyone else, and which can be leveraged in multiple industry sectors.
For instance, Canon's opto-electronics skills enabled it to move from cameras to laser printers to lithographic chip manufacturing equipment.
Honda's engine competency enabled them to expand from motorcycles to cars.
The core competence needed for India is in chip foundries.
Granted, these are enormously expensive at some $10 billion a pop.
India only has two chip foundries, SITAR in Bangalore/Hyderabad and ISRO's Semiconductor Laboratory in Chandigarh.
SCL runs at a 180 nanometers capability, whereas the most uptodate foundries run at around 5 nanometers, several generations ahead.
Yet it's a start.
It was a landmark when IIT Madras announced they had developed 'Shakti', a chip based on open-source RISC-V instruction set architecture from the University of California, Berkeley, and fabricated it at the SCL.
Yet, China has dozens of chip foundries, and we surely we need to build more.
The policy does suggest that 'mega projects' like 'trusted foundries', display, photonics and LED fabrication facilities need to be supported, but also adds that investment in existing foreign foundries might be a good intermediate step.
In my opinion, this isn't enough: A moonshot-like thrust needs to be taken forward to get into chip fabrication in a big way.
On other aspects, the policy takes a conservative approach: Leveraging India's existing competencies.
One is in design, the first half of Electronic System Design and Manufacturing (ESDM).
Fab-less design, the art of designing chips without actually manufacturing them, has been done in India for some time, from the early days of Texas Instruments and Motorola in their Bangalore labs from the mid-1990s.
Britain's ARM Holdings has made a fortune and become the world's leading chip designer (there's a 99% chance that your smartphone has an ARM chip in it) the same way; its licensee Qualcomm and even Apple do the same thing: design chips with outsourced manufacturing.
The second is again in design, but not at the chip level, but at the board-level.
When I ran an incubator for electronics startups, I used to advise them to do their proof of concepts quickly by using open-source hardware like Arduino and Raspberry Pi, which are standard, off-the-shelf products which can be programmed to meet your requirements.
Thus, a quick, working proof of concept for your device can be developed doing almost nothing but software programming.
The next level is to optimise the design of the product using ECAD software, which can quickly get you to a prototype printed circuit board design.
If you have access to a manufacturer of PCBs, and access to mechanical design (for enclosures) and to an assembly line with a supply chain capability, you can go to the next level as well: Which is to be a full-fledged OEM supplier to big brands.
The proposed support for both existing and new Electronic Manufacturing Clusters (EMC) will enhance this capability.
The other area that the policy emphasised at length is support for human resource development as well as Intellectual Property Rights development.
This is a major lacuna, because there aren't enough people being skilled up with the most advanced concepts, nor is there enough R&D leading to strong patent protected rights.
These are systemic problems with the education and R&D cultures of the country, and a lot needs to be done to improve them.
In the final analysis, the policy is adequate in incremental innovation using existing strengths, but it is too timid and doesn't go far enough in articulating a leapfrog vision and a strategic intent.
Nor does it support the development of core competencies that can lead to the disruptive innovations that can help India move to a higher trajectory in the world of advanced technology.
