Technical Experience

My general techincal interests are in controls, stochastic systems, optimization, and machine learning with applications in autonomous and robotic systems.

Industry Experience

Machine Learning in Safety Critical Applications

My industry experience is in the safe application of machine learning in safety critical systems, specifically autonomous vehicles. I have pratical experience in analyzing and verifying machine learning models. I’m particularly interested in the training of models whose resultant attributes are production ready.

Safe Software Architectures

While the autonomy components of a system perform much of the required functions, there is a need for the design of safe software architectures for instances when less than nominal situations are encountered. In such cases, safe software architectures play a pivotal role in handling failures gracefully when the autonomy components of a system encounter an error or in instances when the system enters a domain it is not designed to operate in. I’m interested in the design of safe software architectures that are contract-based with a particular focus on failure mode handling.


Analysis and Control of Stochastic Systems

Stochastic systems are systems that experience or exhibit random behavior. Such systems are ubiquitous in real-world applications such as underwater robotics, resource management, traffic networks, and financial markets, among others. The randomness of stochastic systems makes them challenging to understand, analyze, and control; however, this difficulty highlights the importance of studying such systems.

I’m generally interested in the analysis and control of stochastic systems. On the analysis front, I’m interested in developing computationally efficient tools for the safety verification of stochastic systems. Frequently, it is also important to develop control algorithms that compute control actions that achieve the desired objective with high probability.

Queuing Networks Applied to Charging Networks

Mobility is contemporaneously undergoing seismic changes stemming from evolving technologies that are fundamentally altering the way we transport people and goods. A notable example of such changes is the increased adoption of non-internal combustion engine vehicles, specifically electric vehicles (EVs). This is catalyzed by improved affordability of EVs which is reflected in the conclusions of recent studies that forecast that global new vehicle sales in 2040 will be 58% EVs and the global passenger vehicle market will be 31% electric. The EV revolution promises to transform transportation and mobility while having consequential effects on global energy and transportation infrastructure. With the growing number of EVs, the demands on the EV charging ecosystem (EVCE), i.e., charging facilities, charging schedulers, utility providers, etc., will be greater.

Many aspects of the EVCE are stochastic, i.e., random, in nature, since many of the interactions that occur amongst the entities that comprise the EVCE are nondeterministic. This poses a variety of challenges for both customers and charging facility operators. I’m particularly interested in developing probabilistic techniques for EV charging facility operation using techniques from optimization, control theory, machine learning, and queueing theory.