Arpita Sinha

Current research area

  • On-road path planning of self-driving cars.

  • Optimal graph exploration by multi-agent system

    • for unpredictable patrolling.

    • for mapping.

    • for complete coverage.

    • for 3D structural inspections.

    • for railway track inspections.

  • Application of Space Filling Curves in robotic path planning.

  • Multi-agent pattern generation.

The following are the topic I have worked on:

Exploiting Lissjous curves

Patterns with Unicycle

We see various patterns in nature. Many of them can be expressed mathematically and can be reproduced in computer simulations. What if a robot has to draw it? We design the controller such that an unicycle robot can trace some of these curves. Apart from their aesthetic value, these curves are useful in several civil and military applications like search and surveillance.

  • T. Tripathy and A. Sinha: “Generating Patterns with Unicycles”, to appear in IEEE Transactions on Automatic Control

Target Tracking

In many applications like convoy protection, multiple micro aerial vehicles need to cooperatively track a ground vehicle. The controllers have to be designed taking into account the scalability and robustness issues. However, this is not enough to have a successful operation since there can be problems in communication, sensing and actuations. Some confidence can be achieved by testing on a Hardware-in-loop simulator, which is the ultimate goal of this research. (This work is being carried out jointly with Prof. H. Arya).

  • S. Daingade, A. Sinha, A. Borkar, H. Arya: “A Variant of Cyclic Pursuit for Target Tracking Applications: Theory and Implementatio”, Autonomous Robots, 2015, DOI: 10.1007/s10514-015-9487-3

  • G. Mallik, S, Daingade and A. Sinha: “Scalable Multi-UAV Formation using Consensus based Target Centric Deviated Cyclic Pursuit”, European Journal of Control, 2015, DOI: doi:10.1016/j.ejcon.2015.10.002.

Multi-robot SLAM

Simultaneous localisation and mapping (SLAM) is a challenging problem when there are multiple vehicles exploring the environment simultaneously with limited communication between them. We require the task to be completed in minimum time which means the maps generated by the robots should have minimum overlap. Consequently, merging the maps is an interesting problem. (This work is being carried out jointly with Prof. L. Vachhani)

  • Sarat Chandra N., L. Vachhani and A. Sinha: “Multi-robot Graph Exploration and Map Building with Collision Avoidance: A Decentralized Approach”, Journal of Intelligent and Robotic Systems, 2015, DOI: 10.1007/s10846-015-0309-9

Miscellaneous

  • Satellite formation flying
    Students - Rakesh Warier (PhD, ongoing), Shashank Agarwal (MTech, 2011)

  • Multiagent rendezvous with minimal sensing and course actuation
    Student - Soumya Sahoo (PhD, 2013)

  • Generalized Cyclic Pursuit (my PhD thesis)