In Conversation With Shailesh Kumar

DoJMA: You’re a B.Tech in Mechanical Engineering and an M.S in Computer Science + Mechanical Engineering. What motivated you to turn to innovation in a non-mainstream field like Virtual Reality?

Shailesh: I graduated from IIT Kanpur in 1999. It was the dot-com era. Huge offers were rolled out and I joined a company. For three years, I worked there and was bored by the typical IT work it called for – Java, web-development, etc. I had done mechanical engineering where we dealt with designing, maths and physics, but I wasn’t putting any of that to use. But, I did love computers and programming. While trying to think of an area that could be an amalgamation of programming and engineering, I thought of computer graphics. On researching more about it, I chanced upon Virtual Reality (VR), which seemed to have a lot of potential in it. Though it was only in its budding stage back then, I realised that this stream would grow in its significance with time. I was fascinated and impressed by the Virtual Reality Lab in Iowa State University, which had one of the most advanced projection technologies for VR. I engaged myself in some courses in computer graphics. One of the ‘A’ s I had scored in my undergraduate level studies was in Technical Design. Though many batch-toppers had to struggle to score in Technical Design, it came quite naturally to me. I could imagine and draw pretty easily. So, my design, visualisation skills and my knowledge of mathematics were channeled into computer graphics.

DoJMA: VR/AR still seem to be a little niche. VR is a computationally expensive technology. How do you see it coming to the mainstream consumer? What other challenges are faced by the industry?

Shailesh: The main challenges faced by the industry are the hardware’s form factor and cost. In my talk, I talked about the iPod moment. It happened when you could afford a reasonably good VR headset for $200, which meant that everybody could afford it. Hence, the cost of entry has already been taken care of. Form factor is a problem that still remains. Carrying around a huge headset is inconvenient and putting it on for a reasonable stretch of time is uncomfortable, so the form factor needs to be reduced. HoloLens and Magic Leap are reasonably wearable platforms for AR and people are trying to come up with lighter form factors which you can wear for longer times. Like I mentioned in my talk, maybe in a few years time, it’ll be in the shape of eyewear.
The third challenge faced is the computing platform. Most portable headsets right now have processors which can only handle only a certain kind of 3D data. But as processors become faster, better and more efficient, the process will also become faster and better. They will be able to handle more workload. As a result, the renderings will become more beautiful, more realistic, and the models become bigger and heavier.
The third challenge is what I am working on. If we can offload all the computation from the client to the cloud, then the computation is taken care of. But then, bandwidth becomes the problem. Whatever is rendered there needs to come to my phone in real-time, which creates issues due to limited bandwidth. This issue will potentially be solved by the advent of 5G. This is the entire ecosystem of problems and the potential solutions. For each problem I mentioned, people are coming up with solutions. Companies like Intel and Qualcomm are working on better solutions.

DoJMA: How should college students interested in this field or any emerging field in general prepare themselves?

Shailesh: Nowadays, students are lucky as they have platforms and resources like Coursera, Udemy, Udacity and of course, YouTube. Even I use them to upscale my skills. You don’t have to go hunt for a good book like we had to in my times. Also, these courses are tailor-made. Instead of having big books and figuring out what to read and what not to read, we can go to any of these courses and just start learning.