DST Inspire Faculty Fellow develops algorithms to ensure efficiency of networked systems

Atreyee Kundu, a Department of Science & Technology (DST) INSPIRE Faculty fellow, has designed scheduling and control algorithms for networked control systems whose shared communication networks have a limited communication capacity and are prone to data losses.

The paper written in collaboration with Prof. Daniel Quevedo of Queensland University of Technology, Australia, and accepted for publication in the journal IEEE Transactions on Control of Network Systems studies systems in which multiple devices are accessing a wireless network of limited capacity and prone to loss of information and gives algorithms for the sequence in which the devices can be given access to the network and control inputs that can be applied so that all the devices maintain desired properties to ensure efficiency of the networked system.

Dr. Kundu works on developing new sets of decision and control algorithms to help society take maximum benefits of technology for fixing various wide-ranged problems in various sectors such as health, education, drone flying, and so on. She works on the design of decision and control algorithms for networked control systems. She employs hybrid dynamical systems framework and discrete mathematics as the primary tools for system analysis and algorithm synthesis. Numerically tractable algorithms that she has proposed are particularly useful in setups like smart homes, smart cities, group of robots carrying out a specific task, platoons of autonomous vehicles, and so on.

“In modern-day engineering systems, the process and its control algorithm are often required to be geographically apart. For example, a remote surgery, where the doctor and the patient are at two different locations, a flying drone is being controlled by a ground PC, platoons of autonomous cars plying on the road, etc. The process and the control algorithm talk to each other over a shared wireless communication network, and we call such systems as Networked Control Systems. The real-time feedback control loops present in networked control systems are different from traditional feedback control systems due to communication limitations and uncertainties. Naturally, new sets of decision and control algorithms are a key requirement in modern engineering which is what I design as a DST INSPIRE faculty fellow,” said Dr. Kundu.

Her work is fundamental to a wide class of modern engineering systems and benefits every section of society through a range of areas like healthcare disaster management governance, administration, and so on.

Dr. Kundu has developed stabilization algorithms, scheduling algorithms, and control algorithms for networked control systems that are subject to communication limitations and uncertainties. Her work in the areas of networked control and cyber-physical systems can help making modern engineering facilities like telesurgery, autonomous vehicles, etc., more accessible to everyone in the long run.