Featured Eminent Speaker of ACM India: Aritra Hazra
The ACM India Eminent Speaker Program (ESP) provides ACM Student and Professional chapters and institutional partners in India with direct access to top technology leaders, innovators and researchers who will give talks on contemporary and engaging issues that are important to the computing community. Chapters can invite ESP speakers to give talks as part of events that they hold. ACM India will help cover the costs for travel while the local chapter will arrange accommodation and local logistics. While most of the talks should be in-person to encourage interactions, some of the talks may be done virtually. For more details, see the ACM India ESP Page
Bio. Dr. Aritra Hazra is an Assistant Professor in the Department of Computer Science and Engineering (CSE) at the Indian Institute of Technology (IIT) Kharagpur, from where he also received his PhD in 2015. Prior to this, he served as an Assistant Professor in the Department of CSE at IIT Madras from 2015 to 2017. Dr. Hazra's research interest lies broadly on the areas of Formal Methods, Design Verification, VLSI CAD, Artificial Intelligence and Machine Learning. He has published many research articles in leading journals and well-known conferences, and won two best student paper awards in VLSI Design Conferences (in 2010 and 2017). He also received several accolades for his PhD work, including the ACM (India) Best PhD Dissertation Award 2015 and IESA Technovation–TechnoInventor (PhD) Award 2015. Besides these, Dr. Hazra was a recipient of the INAE (Indian National Academy of Engineering) Young Engineer Award in 2017 and has been a Young Associate of INAE since then. He was also awarded the Associateship of the Indian Academy of Science (IASc) in 2018, the Institute of Engineers India (IEI) Young Engineer Award in 2019, and the Institute Faculty Excellence Award (Assistant Professor category) in 2021. He is a member of IEEE and a professional member of ACM.Please share a little about your research and your lab at IIT Kharagpur.
My research interest broadly lies in the areas of formal methods, design verification, VLSI CAD (computer-aided design), artificial intelligence and machine learning. In our Formal Methods Lab at IIT Kharagpur, we try to address some of the emerging research challenges in control systems, safety-critical hardware and software systems, digital and analog integrated circuits, and secure and intelligent designs. Certifying the safety of such systems is one of the most significant concerns in the leading industries. Formal methods are logic-based foundational techniques originating from artificial intelligence and have been refined over the years through new algorithms and powerful back-end SAT/SMT solvers to prove the correctness of a wide variety of systems. We aim to develop automated frameworks leveraging formal methods to verify and ensure the functional correctness, power management, reliability, security, quality, and robustness attributes of such safety-critical systems. We also assess the vulnerabilities of secure systems, including their resilience to fault and machine learning attacks, and certify them leveraging CAD-based testability and formal learnability analysis frameworks. Further, we also target to explore and devise formal approaches for comprehensive interpretability and ensure robustness of trained expert/learned models so that it cannot be fooled or manipulated by adversarial/malicious perturbations.
How did your career path lead you to research in this area and further join academia?
During my graduation days as a student of Computer Science and Engineering, the handshaking of foundational/logical as well as applied aspects of computing fascinated me to explore further. Later, when I was introduced to the problem of formal verification as part of my post-graduation thesis work, I found it very interesting because of its complexity and theoretical underpinning vis-a-vis practical significance. I continued exploring this domain further in my PhD work and later, where I formally validated the non-functional performance aspects of safety-critical designs, namely their power performance, functional reliability, and security. At present times, with the growth of expert and learned systems, the certification requirements also become more challenging due to their restricted explainability, and hence, formal methods are finding newer research pathways to be enabled here. The theoretical insights of formal methods with such a dynamic and ever-increasingly complex design landscape intrigues me in carrying out research in this area and devising methods that are scalable and practically usable. In addition, I was always passionate about teaching because it made the realization of concepts much clearer and often broadened the thought horizon through interaction -- thereby indirectly helping to identify, formalize, and solve research problems. Hence, inclinations to teaching and engagement in such research automatically paved my way seamlessly into academia.
What are some ways to increase diversity, equity, and inclusion in this area?
In my opinion, involving the students from their graduation time in this domain through various courses and projects will be a starting point. Besides, mentoring workshops and collaboration opportunities with industries in the form of industrial internships can also be beneficial to understanding the specific practical needs in this area. Moreover, carrying out outreach programs involving industry partners will encourage the broader community to engage themselves and gather insights. It is very important to motivate the early participants by showing both the theoretical rigor and practical applicability of the methods being used in this research domain - thereby bridging the gap of foundational understanding with methods working in practice. Finally, the participation needs to be enriched from broader communities of engineering sciences, including computer science, electrical science, and electronics engineering.
What is an important issue you are working on that you think hasn't received enough attention?
I primarily try to showcase how beautiful theories and foundational frameworks in this research domain can be made scalable and practically viable through algorithmic and engineering marvels. Moreover, when one talks about problems related to VLSI (very large-scale integration) or EDA (electronic design automation), or system/chip/semiconductor designs, there is an inherent bias in the thought process that this may be more closely related to electronics or electrical sciences. However, I often try to disseminate that there are computer-aided design (CAD) solutions that are purely mathematical/logical and algorithmic in nature. Through my presentations, I try to bring the diversified participation of various scientific and engineering fields in the growth of this research domain.
What excites you about the topic that you will be talking about as part of the ESP program?
First of all, it is exciting to inform how formal methods can enable practical systems from our day-to-day lives to run with utmost safety, security and reliability. With human-like intelligence built into modern-day systems, it is also overwhelming to realize how theoretical/logical frameworks play a pivotal role in certifying such expert intelligent systems. Moreover, it fascinates me to talk about the growth of artificial intelligence and logic in modern times, surpassing the human cognition capabilities in various disciplines. It is always intriguing to cherish the pathway that led to such advancements of research in a domain, with a hint of where we are leading to and the enormous possibilities that lie in front of us. Personally, it gives me immense pleasure to interact with the audiences having diverse backgrounds and various viewpoints regarding the topics of the talk.
If your ACM Chapter would like to invite Aritra Hazra to give a talk, please follow the guidelines for inviting an ESP Speaker available here.
- Policy Optimization in Reinforcement Learning
- Learning to become Experts in Playing Games
- Formal Methods for Power Intent Verification
You can also start an ACM Chapter or become an ACM Member.
