Research and Projects
Research
Autonomous aerial convoy monitoring using elliptical orbits
Here we propose a vector field based guidance scheme is that makes an aerial agent monitor a convoy of targets moving along a possibly non-linear trajectory on the ground. The scheme first computes a time varying ellipse that encompasses all the targets in the convoy using a simple regression based algorithm. It then ensures convergence of the aerial agent to a trajectory that repeatedly traverses this moving ellipse. The correct tracking behaviour of the scheme is rigorously established along with error bounds, based on the perturbation theory of ordinary differential equations. It is supported by MATLAB simulations and experiments using the AR.Drone 2.0 as the aerial agent and Firebird V robots as the ground convoy in a Vicon motion capture environment.
This strategy is then extended to surveillance of a dynamic ground convoy, moving along a nonlinear trajectory, by team of aerial agents that maintain a uniformly spaced formation on a time-varying elliptical orbit encompassing the convoy. The proposed scheme includes an algorithm for computing feasible elliptical orbits, an updated vector field guidance law for agent motion along the desired orbit, and a cooperative strategy to control the speeds of the aerial agents in order to quickly achieve and maintain the desired formation. The proposed scheme achieves mission objectives while accounting for the linear and angular speed constraints of the aerial agents. The scheme is validated through simulations and actual experiments, with ground robots as the convoy, and a team of Crazyflie 2.0 quadrotors as the aerial agents in a motion capture environment.
Related publications:
A. V. Borkar, V. S. Borkar and A. Sinha, "Vector Field Guidance for Convoy Monitoring Using Elliptical Orbits", 56th IEEE Conference on Decision and Control (CDC), Melbourne, Australia, December 2017, Pages 918-924.
A. V. Borkar, V. S. Borkar and A. Sinha, "Aerial Monitoring of Slow Moving Convoys using Elliptical Orbits", European Journal of Control, Volume 46, March 2019, Pages 90-102.
A. V. Borkar and G. Chowdhary, "Multi-agent Aerial Monitoring of Moving Convoys using Elliptical Orbits," 2021 IEEE International Conference on Robotics and Automation (ICRA), 2021, pp. 8999-9005.
Collision-free multi-agent formation on Lissajous curves for area surveillance.
Here we propose a multi-agent formation that moves on a Lissajous curve in a collision-free manner. This multi-agent formation simultaneously performs multiple surveillance tasks such as repeated collision-free area surveillance, guaranteed complete sensor coverage of the rectangular area, and finite time target detection and entrapping. The agents are considered have finite non-zero dimensions and a sufficient upper bound on the agent dimensions is derived to guarantee collision free motion of the formation on the Lissajous curve. We have also proposed an algorithm to select the number of agents to be deployed, and the optimal Lissajous curve to be used (in terms of area coverage time and bound on agent size). This strategy is easily scalable and doesn't need any cooperation among the agents as collision-free motion is achieved simply by leveraging the properties of the Lissajous curve. This strategy was validated both in MATLAB simulations and experiments with Firebird V robots.
Related publication: A. Borkar, A. Sinha, L. Vachhani, and H. Arya, "Collision-free trajectory planning on Lissajous curves for repeated multi-agent coverage and target detection", Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Daejeon, Korea, October 2016, Pages 1417-1422.
Reconfigurable quadrotor formations on Lissajous curves for area surveillance
We extend the above idea of a non-cooperating multi-agent formation on a Lissajous curve to a cooperating and reconfigurable multi-quadrotor formation on Lissajous curves for collision-free aerial surveillance of a rectangular area. The reconfiguration strategies are proposed for quadrotor addition, removal or replaced using a decentralized cooperating scheme. We have validated our results through MATLAB simulations for agents having a non-zero size satisfying a theoretically derived size bound. To demonstrate the practical applicability of the proposed surveillance and reconfiguration strategies, we have performed simulations for quadrotors in a ROS-Gazebo based Software-In-The-Loop (SITL) simulator and have implemented the same with a team of five indigenously developed quadrotors operating in a Vicon motion capture environment.
Related publication: A. V. Borkar, S. Hangal, H. Arya, A. Sinha and L. Vachhani, "Reconfigurable formations of quadrotors on Lissajous curves for surveillance applications", European Journal of Control, Volume 56, November 2020, Pages 274-288.
Application of Lissajous curves in trajectory planning of multiple agents.
Lissajous curves have been used in various engineering applications such as optics, imaging, antenna scan, machining, as well as mobile robotics. We propose and analytically justify a Lissajous curve based trajectory planning strategy for aerial multi-agent systems to achieve the following objectives simultaneously:
Collision free paths for repeated coverage of a region while maintaining a closed sensor ring around a specified center for all time.
Guaranteed detection of any stationary or moving object enclosed within the ring in finite time without the possibility of undetected escape.
This leverages known and some novel properties of Lissajous curves that we establish as a part of this work. This has several potential applications in civil and military missions such as search and surveillance, repeated patrolling, target detection and capture, and the proposed strategy meets all these objectives simultaneously. We validate the proposed strategy through simulations and experiments using differential drive ground robots.
Related publication: A. V. Borkar, A. Sinha, L. Vachhani and H. Arya, "Application of Lissajous curves in trajectory planning of multiple agents", Autonomous Robots, Volume 44, No. 2, 2020, Pages 233-250.
Projects
(In collaboration with Sri Theja Vuppala)
We built a DJI Flamewheel 450 quadrotor, equipped with a Pixhawk 4 autopilot interfaced to a NVIDIA Jetson Nano companion computer. The system was set up to operate in the default flight modes of the Pixhawk flight stack as well as in the offboard flight mode for autonomous operations. The drone was flown to demonstrate autonomous takeoff, waypoint navigation and autonomous landing.
3D SLAM using Realsense L515 and RTABMAP
The RTABMAP 3D SLAM algorithm was deployed on a Earthsense Terrasentia robot using the Intel Realsense L515 camera and an Intel NUC for the localization and Mapping computations. This implementation does not rely on any wheel odometry as the camera was mounted on a swiveling joint which could rotate around the vertical axis. A demonstration of indoor localization and mapping is shown in the video.
Research & Development in collaborative missions Project code: 11DST027 Supervisor: Prof. Hemendra Arya
(In collaboration with Dr. Sangeeta Daingade, Prof. Arpita Sinha, Srishti Tiwari and Satyaswaroop Yadati)
This was a Department of Science and Technology (DST) sponsored undertaking in the Department of Aerospace Engineering, IIT Bombay. We worked primarily towards addressing issues related to implementation of cooperative missions on Miniature Aerial Vehicles with simulations on a Hardware-In-The-Loop Simulator (HILS) for simulating cooperative multi-MAV missions. The work done in this project included:
Developing real time monitoring capability of flight parameters for each MAV.
Implementing attitude estimation algorithms in the HILS simulations for MAVs.
Incorporating realistic wind models (e.g., Dryden Model) in simulation.
Simulating variants of the cyclic pursuit algorithms in the HILS system.
Related publications:
S. Daingade, A. Sinha, A. V. Borkar and H. Arya, "A Variant of Cyclic Pursuit for Target Tracking Applications: Theory and Implementation", Autonomous Robots, Volume 40, No. 4, 2016, Pages 669-686.
S. Daingade, A. Sinha, A. Borkar and H. Arya, "Multi UAV formation Control for Target Tracking", Proceedings of Indian Control Conference (ICC), Chennai, India, January 2015, Pages 25-30.
S. Daingade, A. Borkar, A. Sinha, and H. Arya, "Study of Target Centric Cyclic Pursuit for MAVs using Hardware In Loop Simulator", Proceedings of AIAA Modeling and Simulation Technologies Conference, Maryland, USA, January 2014, Page 1345.
S. Daingade, A. Borkar, A. Sinha and H. Arya, "Implementation of Collective Target Enclosing Strategy with Multiple MAVs using Hardware In Loop Simulator", Proceedings of International Conference on Intelligent Unmanned Systems (ICIUS), Jaipur, September 2013, Volume 9.
A. V. Borkar, D. Krishnan, S. Tiwari and H. Arya, "Effects of Wind on Cooperative Missions", Proceedings of International Conference on Recent Advances in Design, Development and Operation of Micro Aerial Vehicles, Hyderabad, December 2012.
S. Yadati, A. Borkar and H. Arya, "Evaluation of Attitude Determination Algorithms using Hardware-In-Loop-Simulator", Proceedings of International Conference on Recent Advance in Design, Development and Operation of Micro Aerial Vehicles, Hyderabad, December 2012.
Projects as a student
Master of Technology Dissertation
Cooperative Control for Fixed Wing Miniature Aerial Vehicles (MAVs) Supervisors: Prof. Hemendra Arya and Prof. K. Sudhakar
(In collaboration with Dileep Krishnan, Prasanna Shevare, Prateek Jolly and Anand Biradar)
We implemented a Time Division Multiplexing (TDM) based communication protocol for the wireless network between multiple MAVs and the ground station for effective cooperation. We upgraded the existing ground station software for in-flight autopilot gain tuning of individual MAVs and for receiving and displaying telemetry data from upto eight MAVs for easy monitoring of the multi-MAV mission. We also developed a MATLAB based Hardware In the Loop Simulator (HILS) which does a real time simulation of a 6-DOF aircraft model for each MAV. This simulator has the flight controller and the communication modules as the Hardware-In-The-Loop for real time simulation of cooperative missions of upto eight MAVs. Using this setup we first validated the flight code and the communications protocols by implementing different cooperative missions: orbit tracking, leader-follower, way-point navigation, and different variants of the cyclic pursuit laws. Flight testing of some of these missions was done with three fixed wing MAVs (small UAVs with a 1 meter wingspan). Videos of some of the flight tests and some HILS simulations can be found below.
Related publications:
D. Krishnan, A. Borkar and H. Arya, "An Elegant Hardware in Loop Simulator for Cooperative Missions of MAVs", Proceedings of AIAA Infotech@Aerospace 2012, Garden Grove, California, June 2012, Page 2419.
D. Krishnan, A. Borkar, P. Shevare and H. Arya, "Hardware in Loop Simulator for Cooperative Missions", Proceedings of International Conference on Advances in Control and Optimization of Dynamical Systems (ACODS), Bengaluru, February 2012.
Bachelor of Engineering Dissertation
Here we have constructed an autonomous wheeled robot equipped with sensors for detecting fires and obstacles. The robot operates in an apartment where there are infra-red sensors fitted in each room to detect fires. When a fire is detected in a room the sensors alert the robot via a wireless communication link. The robot is equipped with proximity sensors and it uses a simple wall following strategy to navigate through the model apartment. A simple counter based algorithm is used to skip the rooms where there is no fire detected. Since this was only a proof of concept, candles were used to simulate fires and a CPU fan was used to put them out when detected as seen in the video below.
Project Group: Aseem V Borkar, Devdatt Haldipur, Siddhesh Bhosale and Renuka Sunder
ROS based spherical robot
This robot is controlled by a ROS node written in python running in a ROS Kinetic environment on a Raspberry Pi zero W. The forward and reverse motion is performed using two 100 RPM geared DC motors, and direction is controlled by tilting the sphere. This is done by shifting center of gravity of the robot, by swinging a pendulum with brass weights at the bottom, using MG-995 high torque servo motors. The spherical shell used is an exercise ball for small animals, available at most pet shops. To reduce wobbling and oscillations, an active damping control is implemented which uses orientation feedback from the 9-axis MPU 9250 sensor.
ROS based differential drive robot with pan-tilt camera control
This robot is controlled using the ROS Kinetic environment on a Raspberry Pi 3B (initial version with a Raspberry Pi Zero W). Each wheel is rotated by a 60 RPM N-20 geared DC motor, and the motors are actuated to implement the differential drive scheme using the L-298N motor-driver. The Raspberry Pi camera is mounted on light weight pan-tilt platform actuated by two Tower Pro 9 gm servomotors. The ROS node written in python allows for both autonomous and manual control with a computer joystick for locomotion as well as camera orientation. The live camera feed can be viewed remotely on ROS topics for further processing.
Trajectory tracking on a ball and plate control system
The experiment was performed on the amazing ball setup at the Dynamics and Control Lab, Dept. of Aerospace Engineering IIT Bombay.
The setup has two servo motors controlling the pitch and roll of the flat plate. The flat plate is touch-sensitive and gives the position coordinates of the metal ball as feedback to the micro-controller. For this experiment the plate was actuated to move the ball along a circle, and a Lissajous curve.
Line following robots
This is a very simple line following robot that follows a black line on a white surface. The circuit doesn't use a micro-controller and is very simple, using only two ICs: a LM-339 Quad comparator and a L293D Motor Driver. For sensing the line two CNY-70 optical sensor are used, each of which consists of a photo transistor and IR LED pair. The robot is powered by a 6V rechargeable lead acid battery.
Here the Firebird III robot is used to perform a line following task. The robot has an ATMEGA 128 micro-controller and line sensors. The robot is programmed to follow lines and select path branches. It completes the course once and then stops in the rectangle where a box is kept. It senses the box using a proximity sensor. This was done as a project for the course AE-773 Applied Mechatronics at the Dept. of Aerospace Engineering, IIT Bombay.