Research and Technical Projects

Major research activities and technical projects from my undergraduate and graduate studies

Decentralized Multi-Agent Interaction with Tumbling Objects in Orbit

A John-Sabu, BT Lopez

Abstract: This research involves the investigation into techniques for the interaction of multiple chaser spacecraft with uncontrollable (and sometimes limitedly observable) tumbling objects in orbit. This includes the implementation of robust tube-based model predictive control for trajectory tracking and planning, improvised auction algorithms for task allocation, and the application of orbital dynamics in the target-centered coordinate frame.

A John-Sabu, D Mukherjee

Abstract: It is well acknowledged that human-made technology is not always at par with human curiosity, and an example is the inability to send large telescopes to outer space despite their higher resolution and less atmospheric interference. In this paper, we develop a framework for autonomous in-orbit construction using spacecraft formation such that a large telescope can be built in an elliptic orbit using multiple spacecraft. We split this problem into four steps for converging the position and attitude of each spacecraft at predefined values around a central spacecraft. Each spacecraft performs attitude synchronization with its neighbors to match its three degrees of freedom in orientation as a parabolic mirror. Simulations validate our proposed methods and the paper concludes with an open possibility of using other techniques to improve upon existing results.

Achieving the 'Unorganized Pre-assembly' stage via the Distributed Auction Algorithm

A John-Sabu, D Mukherjee

Abstract: The need to match the advancements in communication and transportation calls for decentralized formation algorithms between satellites since controlling them from ground stations on earth fails due to communication lags and interruptions. It is also required that the satellites are capable of aligning themselves based on the collective decisions made by the group. As a result, each satellite requires an attitude control mechanism blended with the formation control algorithm to control its heading and angular velocities. We look into studies in these topics both individually and collectively. We also discuss attitude control and maneuver strategies from a renowned textbook in the field. Further, we explain results from the studies along with possible scopes of improvement. Finally, we provide a picture of how research on this field can move forward in the future with the ideas obtained from the above sources.

R Ramesh, A John-Sabu, Harshitha S, S Ramesh, V Navada B, M Arunachalam, B Amrutur

Abstract: Modern requirements for autonomous drones require the constant determination of the location of the drone. Ultra Wide Band (UWB) radios fulfill such requirements and offer high precision localization and fast position update rates at lower costs and battery consumption when compared to lidars, while also having greater network availability than GPS in a densely forested campus or an indoor setting. In this research, we develop a novel protocol to localize drones for such applications using a Time Difference of Arrival (TDoA) approach. We present the results obtained from experimentation in an indoor environment in the midst of other facilities such as WiFi or mobile network coverage, and reason the scalability of the algorithm from the decrease in information exchanges for each cycle of localization. A linear Kalman filter is also incorporated to eliminate clock drift by estimating clock skews and hence the time of arrival of ranging messages.

Exploiting structural instabilities to build
Logic Gates, Neural Networks, and beyond

A Prototype for the Logic Gate

A John-Sabu, A Ramabathiran

Abstract: We utilize the structural instability of buckling to physically emulate digital circuits. We aim at its constructive implementation by differentiating the two buckled states of each unit cell as HIGH or LOW which is propagated throughout the network. At suitable points, logic gates (AND and NOT) are implemented, and the mechanical flow is maintained using manipulators such as dimensional shifters and splitters. We look forward to redefining the digital binary network as a binary artificial neural network wherein multiple inputs are accepted by neurons that replace logic gates.

Abstract: The IIT Bombay Student Satellite Program is a landmark project taken up by IIT Bombay students with the vision to make the institute an internationally acclaimed center of excellence in satellite and space technology. As an interdisciplinary team of over 50 students from across the institute, we are engaged in projects ranging from the design of space systems of various form factors to the development of custom payload. Following the launch of our first satellite Pratham with a payload to calculate ionospheric Total Electron Count (TEC), we have been developing our second satellite, Advitiy (literally meaning 'second to none' or 'unparalleled'), the lightest actively-controlled 1U CubeSat, with the objective of transmitting stored and uplinked images via SSTV. As part of this project, we have developed the design of the antenna deployment system as well as a closed-loop simulation framework to aid the attitude determination and controls subsystem. At the moment, following the announcement of the opportunity for the PSLV Stage 4 Orbital Platform by ISRO, we have carried forward the development of suitable CubeSat subsystems with relevance to the indigenous research community following a halt on the work on Advitiy.

Real-time Detection and Localization of Ground Objects
using a Fleet of Unmanned Aerial Vehicles

D Chaturvedi, A Rahar, A John-Sabu, A Malekar, A Agarwal, J Prashath

Abstract: The search and rescue of military personnel in hostile environments bring in the need for faster, robust, and less detectable systems. Such systems can also prove helpful in disaster and emergency analysis, as well as for reconnaissance missions. We develop an aerial multi-robot system that traverses across a grassy field (40m x 40m) and detects up to 5 green cubes of length 15cm. This fleet of unmanned aerial vehicles (UAVs) takes flight and lands autonomously. Each drone communicates the target location to others in a decentralized manner and plans its path based on the same.

Control and ROS integration segments in individual drones

N Randad, A John-Sabu, V Bhagyawant, J Prashath, Y Patil, K Patil

Abstract: North-East India has several favorable features for agriculture from fertile land to abundant water resources (high and dependable rainfall). It has the potential to utilize these optimally for agricultural productivity. We propose the design of a cost-effective, eco-friendly, and easily operable agricultural robot to convert the incumbent but less effective subsistence farming into efficient mechanized farming by harnessing appropriate niche potentials of marginal mountain land. Our Agrobot deals in particular with effective braking in hilly terrains, transplantation of plants, seed metering, and weed removal, and we do suggest extra features that may supplement the needs of the farmer on an as-needed basis.

Find a non-exhaustive collection of reports from course projects and relevant assignments: Publications and Reports