Cost of connecting any device to internet would be sub-$1; so IoT is going to get bigger: Nambirajan Seshadri of Broadcom
Nambirajan Seshadri, who graduated from Regional Engineering College, Trichy, before moving to the US to work at AT&T, joined Broadcom in 1999 and has gone on to file more than 75 patents in networking and chip design. As senior vice president and CTO of Broadband & Connectivity Group at Broadcom, the over $8 billion dollar fabless semiconductor company, he is building an ecosystem of ideas in the Internet of Things (IoT) space. He spoke to Techcircle.in on opportunities in IoT and the areas which hold the most potential. Edited excerpts:
Can you share your thoughts on the potential of exploiting IoT for businesses? Which segments will tap into the opportunity?
IoT is in its infancy. It is very small. There are talks that there will be 50 billion connected devices by 2020. Cell phones is a big piece of it. Many things are already connected and by the time we are done, there might be 10 billion devices that are connected in a year or so. For example, there is no reason any toy you buy shouldn't be connected to the internet. The cost of connectivity will be a dollar or less than a dollar, and anything that maybe $15 or $20 should be connected to the internet. What value you get from that is a different question. I clearly see that happening.
Manufacturing would be a good place for IoT. Healthcare would be a big place. Wearables is big; not just watches but clothes too. When I go to kirana shops and retail stores, I see enormous opportunities. In health, there are interesting pilots. A wearable device that was highlighted in a conference paper published from Singapore is an example. It is a device that you wear on your wrist. When you eat, it captures the number of movements you do to tell the number of bites. Also, when you drink water, you do an entirely different movement which it tracks. All of these point to a radical opportunity. The cost of what you do for $100 will come down because what you are doing today, you can do with the wireless internet connectivity for embedded devices (WICED) model for $20 tomorrow.
How would Broadcom's WICED be exploited by startups?
We are trying to answer how a small-time manufacturer can build an internet connection. This is where concepts like WICED—a development system or kit on which secure embedded networking application systems can be built—comes to play to help anyone wanting to connect to the internet. How does the ecosystem grow? How does IoT shape up and how is it going to be used? There will be different verticals that startups will work on.
There is a lot of noise around IoT startups but we are yet to see a big success, at least from India. How will the ecosystem for startups tap into IoT?
In the global world of startups, the money is moving towards social media and apps—limited amount is poured into semiconductors. Little more than that is going into system companies who are building interesting products. But, unlike the past, when building products was not funded, we are beginning to see the rapid ability to develop prototypes. People are able to build prototypes due to open source and availability of money to build prototypes and samples. The lifetime is getting shorter and the money it takes is getting smaller. Because of this, we are seeing interesting product companies. Tremendous amount of money is going into software and that will continue.
The opportunity for a country like India is relatable. The app play, products and data analytics are great areas for India. It is a unique case from the traffic point of view. Someone can build a cheap broadband infrastructure based on standard pieces of standardised silicon whether it is Wi-Fi, 60 GHz or cellular. From the infrastructure point of view, we are beginning to have software defined networks and networking platforms. So I think the opportunity exists in a place like India to create very interesting products that are unique to the environment. Twenty years ago I would not have said that.
Again, without naming names, people were designing products tailor made to the Indian market, adopting mobile technologies for fixed wireless. The world of mobile is global and it will drive things to get cheaper. You don't need to do anything special. Today the world has become extremely standardised and we are building things that are extremely cheap. We can now use that to go after new markets that do not require a whole lot of expense. In the US context, any startup can now get $100,000 to $200,000 with a good story. So VCs would probably fund 10 of those—spending say a million dollars. Of the 10 companies, they will choose one or two into which they will invest next million dollars each.
It is essentially being able to rapidly prototype and figure out if there is market success traction or not. If not, just kill it and keep going. It can be done with very interesting products.
The basic infrastructure has to grow. The VCs have a massive role to play. But only if they can divert some attention to make in India products.
(Edited by Joby Puthuparampil Johnson)