ABAnand Balakrishnan

Anand Balakrishnan

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Anand Balakrishnan has a total of 111 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Monitoring Spatially Distributed Cyber-Physical Systems with Alternating Finite Automata

    by Anand Balakrishnan, Sheryl Paul, Simone Silvetti, Laura Nenzi, Jyotirmoy V. Deshmukh


  • Open Access

    Multi-agent Path Finding for Timed Tasks using Evolutionary Games

    by Sheryl Paul, Anand Balakrishnan, Xin Qin, Jyotirmoy V. Deshmukh

    Autonomous multi-agent systems such as hospital robots and package delivery drones often operate in highly uncertain environments and are expected to achieve complex temporal task objectives while ensuring safety. While learning-based methods such as reinforcement learning are popular methods to train single and multi-agent autonomous systems under user-specified and state-based reward functions, applying these methods to satisfy trajectory-level task objectives is a challenging problem. Our first contribution is the use of weighted automata to specify trajectory-level objectives, such that, maximal paths induced in the weighted automaton correspond to desired trajectory-level behaviors. We show how weighted automata-based specifications go beyond timeliness properties focused on deadlines to performance properties such as expeditiousness. Our second contribution is the use of evolutionary game theory (EGT) principles to train homogeneous multi-agent teams targeting homogeneous task objectives. We show how shared experiences of agents and EGT-based policy updates allow us to outperform state-of-the-art reinforcement learning (RL) methods in minimizing path length by nearly 30\% in large spaces. We also show that our algorithm is computationally faster than deep RL methods by at least an order of magnitude. Additionally our results indicate that it scales better with an increase in the number of agents as compared to other methods.

  • Open Access

    Motion Planning for Automata-based Objectives using Efficient Gradient-based Methods

    by Anand Balakrishnan, Merve Atasever, Jyotirmoy V. Deshmukh

    In recent years, there has been increasing interest in using formal methods-based techniques to safely achieve temporal tasks, such as timed sequence of goals, or patrolling objectives. Such tasks are often expressed in real-time logics such as Signal Temporal Logic (STL), whereby, the logical specification is encoded into an optimization problem. Such approaches usually involve optimizing over the quantitative semantics, or robustness degree, of the logic over bounded horizons: the semantics can be encoded as mixed-integer linear constraints or into smooth approximations of the robustness degree. A major limitation of this approach is that it faces scalability challenges with respect to temporal complexity: for example, encoding long-term tasks requires storing the entire history of the system. In this paper, we present a quantitative generalization of such tasks in the form of symbolic automata objectives. Specifically, we show that symbolic automata can be expressed as matrix operators that lend themselves to automatic differentiation, allowing for the use of off-the-shelf gradient-based optimizers. We show how this helps solve the need to store arbitrarily long system trajectories, while efficiently leveraging the task structure encoded in the automaton.

  • Multi-agent Path Finding for Timed Tasks Using Evolutionary Games

    by Sheryl Paul, Anand Balakrishnan, Xin Qin, Jyotirmoy V. Deshmukh


  • Open Access

    The small molecule inhibitor of SARS-CoV-2 3CLpro EDP-235 prevents viral replication and transmission in vivo

    by Michael H. J. Rhodin, Archie C. Reyes, Anand Balakrishnan, Nalini Bisht, Nicole M. Kelly, Joyce Sweeney Gibbons, Jonathan Lloyd, Michael Vaine, Tessa Cressey, Miranda Crepeau, et al.

    The COVID-19 pandemic has led to the deaths of millions of people and severe global economic impacts. Small molecule therapeutics have played an important role in the fight against SARS-CoV-2, the virus responsible for COVID-19, but their efficacy has been limited in scope and availability, with many people unable to access their benefits, and better options are needed. EDP-235 is specifically designed to inhibit the SARS-CoV-2 3CLpro, with potent nanomolar activity against all SARS-CoV-2 variants to date, as well as clinically relevant human and zoonotic coronaviruses. EDP-235 maintains potency against variants bearing mutations associated with nirmatrelvir resistance. Additionally, EDP-235 demonstrates a ≥ 500-fold selectivity index against multiple host proteases. In a male Syrian hamster model of COVID-19, EDP-235 suppresses SARS-CoV-2 replication and viral-induced hamster lung pathology. In a female ferret model, EDP-235 inhibits production of SARS-CoV-2 infectious virus and RNA at multiple anatomical sites. Furthermore, SARS-CoV-2 contact transmission does not occur when naïve ferrets are co-housed with infected, EDP-235-treated ferrets. Collectively, these results demonstrate that EDP-235 is a broad-spectrum coronavirus inhibitor with efficacy in animal models of primary infection and transmission.


  • Safety Monitoring for Pedestrian Detection in Adverse Conditions

    by Swapnil Mallick, Shuvam Ghosal, Anand Balakrishnan, Jyotirmoy Deshmukh


  • Open Access

    Poster Abstract: Model-Free Reinforcement Learning for Symbolic Automata-encoded Objectives | 25th ACM International Conference on Hybrid Systems: Computation and Control

    by Anand Balakrishnan, Stefan Jaksic, Edgar Aguilar, Dejan Nickovic, Jyotirmoy Deshmukh

    In this work, we propose the use of symbolic automata as formal specifications for reinforcement learning agents. The use of symbolic automata serves as a generalization of both bounded-time temporal logic-based specifications and deterministic finite automata, allowing us to describe input alphabets over metric spaces. Furthermore, our use of symbolic automata allows us to define non-sparse potential-based rewards which empirically shape the reward surface, leading to better convergence during RL. We also show that our potential-based rewarding strategy still allows us to obtain the policy that maximizes the satisfaction of the given specification.


  • Open Access

    Model-Free Reinforcement Learning for Symbolic Automata-encoded Objectives

    by Anand Balakrishnan, Stefan Jakšić, Edgar A. Aguilar, Dejan Ničković, Jyotirmoy V. Deshmukh

    Reinforcement learning (RL) is a popular approach for robotic path planning in uncertain environments. However, the control policies trained for an RL agent crucially depend on user-defined, state-based reward functions. Poorly designed rewards can lead to policies that do get maximal rewards but fail to satisfy desired task objectives or are unsafe. There are several examples of the use of formal languages such as temporal logics and automata to specify high-level task specifications for robots (in lieu of Markovian rewards). Recent efforts have focused on inferring state-based rewards from formal specifications; here, the goal is to provide (probabilistic) guarantees that the policy learned using RL (with the inferred rewards) satisfies the high-level formal specification. A key drawback of several of these techniques is that the rewards that they infer are sparse: the agent receives positive rewards only upon completion of the task and no rewards otherwise. This naturally leads to poor convergence properties and high variance during RL. In this work, we propose using formal specifications in the form of symbolic automata: these serve as a generalization of both bounded-time temporal logic-based specifications as well as automata. Furthermore, our use of symbolic automata allows us to define non-sparse potential-based rewards which empirically shape the reward surface, leading to better convergence during RL. We also show that our potential-based rewarding strategy still allows us to obtain the policy that maximizes the satisfaction of the given specification.

  • PerceMon: Online Monitoring for Perception Systems

    by Anand Balakrishnan, Jyotirmoy Deshmukh, Bardh Hoxha, Tomoya Yamaguchi, Georgios Fainekos


  • Biochemical Characterization of Respiratory Syncytial Virus RNA-Dependent RNA Polymerase Complex

    by Anand Balakrishnan, Edmund Price, Catherine Luu, Jacob Shaul, Charles Wartchow, John Cantwell, Todd Michael Vo, Michael DiDonato, Glen Spraggon, Mohammad Hekmat-Nejad

    RNA-dependent RNA polymerases (RdRP) from non-segmented negative strand (NNS) RNA viruses perform both mRNA transcription and genome replication and these activities are regulated by their interactions with RNA and other accessory proteins within the ribonucleoprotein (RNP) complex. Detailed biochemical characterization of these enzymatic activities and their regulation is essential for understanding the life cycles of many pathogenic RNA viruses and for antiviral drug discovery. We developed biochemical and biophysical kinetic methods to study the RNA synthesis and RNA binding activities of respiratory syncytial virus (RSV) L/P RdRP. We determined that the intact L protein is essential for RdRP activity and in truncated L protein constructs RdRP activity is abrogated due to their deficiency in RNA template binding. These results are in agreement with the observation of an RNA template-binding tunnel at the interface of RdRP and capping domains in RSV and vesicular stomatitis virus (VSV) L protein cryo-EM structures. We also describe non-radiometric assays for measuring RNA binding and RNA polymerization activity of RSV RdRP which are amenable to compound screening and profiling.