Arpan Pal, a distinguished Chief Scientist and Research Area Head at TCS Research, has carved a prominent niche in Intelligent Sensing, Signal Processing & AI, and Edge Computing. With an impressive career spanning over three decades, he has contributed significantly to the advancements in Embedded Devices and Intelligent Systems. His expertise lies at the intersection of hardware and software, where he thrives, making significant contributions to embedded systems.

In this interview, Arpan delves into the intricacies of his career journey, shedding light on the inspirations that led him to pursue a path in embedded systems and the subsequent evolution of his expectations. Furthermore, he generously shares insights into the surprises and challenges encountered, emphasizing the critical balance between technological innovation and real-world applications. As a seasoned professional, Arpan offers invaluable advice for aspiring scientists and engineers in the field, providing a roadmap for success that revolves around a deep understanding of hardware-software co-design and the adaptability to emerging technologies.

Arpan envisions a future for Embedded Systems that is deeply rooted in the principles of power efficiency and sustainable computing. With his visionary perspective on the transformative potential of brain-inspired neuromorphic computing and spiking neural networks, Arpan anticipates a paradigm shift towards energy-conscious AI systems. Pal’s remarkable contributions guide the future of technology and innovation as we delve deeper into the world of embedded intelligent systems.

What inspired you to go into embedded systems? Would you say your career has matched what your original expectations were? If so, what? If not, why not?


For me, embedded systems had been a natural fit since I liked both hardware and programming and my first two jobs were in hardware-driven embedded systems – one was to build a microcontroller-based PSTN telephone exchange monitoring system, and the other was to build missile seeker systems. As I began working on these projects, I realized that I loved embedded systems because it is the only field that allows one to work at the intersection of hardware and software and requires knowledge of both electronics and software programming.

So far, the experience and the journey have exceeded my expectations. The biggest satisfaction is to see how embedded systems are making a comeback in the form of Edge Computing and how the concept of “AI at the Edge” is becoming mainstream for IoT, Robotics and AR/VR as it enables reliable, low latency, low power yet privacy-preserving intelligence at the edge.

What in your career has surprised you the most? Are there any challenges you overcame that you’d like to share?


The biggest surprise in the early part of my career was to discover the possibility, for a given use case, of building a computationally lighter version of a sophisticated complex algorithm when faced with the compute/memory/power constraint of embedded systems without any significant compromise on the algorithm performance and I have applied this understanding again and again in my career.

The main challenge in embedded systems research is how to marry technological novelty to a visible and useful impact in the application. When I worked in missile technology, this challenge manifested itself in designing novel real-time target-tracking algorithms that can run on a DSP chip. In my sensing work in TCS for healthcare, this meant designing AIML/Signal Processing algorithms that consume as little power as possible so that they can work with wearables. Our Industry 4.0 intelligent IoT work involved designing systems that provide real-time or near-real-time response with deterministic latency.

The other challenge is at the platform level, where we have come a long way from tiny microcontrollers to DSP processors to AI chipset accelerators. But what has not changed is that an algorithm will always need more time, memory, and power than is available in the target embedded hardware – optimizing it to fit the target hardware is always a challenging task that requires embedded engineering expertise.

What resources or skills did you find most helpful when moving up in your career?


Key skills are as follows:

  • A thorough understanding of hardware system features and limitations is essential for abstracting their implications for embedded applications.
  • When dealing with real-time systems, how to make software optimally utilize the hardware – hardware-software co-design is the key.
  • Understanding on how to map the impact of an application to the novelty of an embedded system in terms of system/technology, and how an application-level constraint will translate into system-level constraint in an embedded system.

What advice would you give to scientists/engineers in embedded systems?


The first piece of advice will be to understand the beauty and the nuances of the hardware-software co-design in embedded systems, which is unique in terms of hardware capability and software features.

The second piece of advice will be to keep an open mind and be ready to adapt to new technologies/techniques as they come. Let’s take an example – In today’s world AI is the hype word; however, AI on embedded systems is not really well-understood yet. Embedded Edge Computing technology is coming up in a big way to address this.

The third advice is to identify a problem and then use technology to solve it, rather than going bottom-up to build a novel technology system first and then look for its suitable application.

What do you see as the future of Embedded Systems?


When will embedded intelligent systems become truly power-aware? Green computing is indispensable as we forge towards a sustainable future. Embedded System engineers are inherently trained to make their algorithms work on low-power, low-latency-constrained embedded devices. The same principles need to be applied to transform over-parameterized ultra-large and power-hungry AI models into power-efficient AI systems.

Our brain computation needs only 20 Watts, while a typical GPU cluster may need tens of kilowatts of power – how do we design AI systems that consume power in the order of our brain? In the area of low-power embedded systems, brain-inspired neuromorphic computing and spiking neural networks (SNN) tailor-made for non-Von-Neumann Neuromorphic architecture will result in significant power saving. SNN on Neuromorphic architecture is a great example of nature-inspired hardware-software co-design.

Learn More About Arpan Pal


Infographic of Arpan Pal's career timeline.

Arpan Pal has more than 30 years of experience in the areas of Intelligent Sensing, Signal Processing &AI, Edge Computing and

Affective Computing. Currently, as Distinguished Chief Scientist and Research Area Head, Embedded Devices and Intelligent Systems, TCS Research, he is working in the areas of Connected Health, Smart Manufacturing, Smart Retail and Remote Sensing.

Arpan has been on the editorial board of notable journals like ACM Transactions on Embedded Systems, and Springer Nature Journal on

 Computer Science. Additionally, he is on the TPC of notable conferences like IEEE Sensors, ICASSP, and EUSIPCO. He has filed 180+ patents (out of which 95+ were granted in different geographies) and have published 160+ papers and book chapters in reputed conferences and journals. He has also written three complete books on IoT, Digital Twins in Manufacturing, and Application AI in Cardiac screening. He is on the governing/review/advisory board of some Indian Government organizations like CSIR, and MeitY, as well as of educational Institutions like IIT, IIIT, and Technology Innovation Hubs. Arpan is two times winner of the Tata Group top Innovation award in Tata InnoVista under Piloted technology category.

Prior to joining Tata Consultancy Services (TCS), Arpan had worked for DRDO, India as Scientist for Missile Seeker Systems and in Rebeca Technologies as their Head of Real-time Systems. He has a B. Tech and M. Tech degree from IIT, Kharagpur, India and PhD. from Aalborg University, Denmark.