Research Interests
A summary of research activities being carried out in our group can be viewed here. Ongoing ProjectsVibration Tolerant Robust Controller Design for Gun Mounted on Naval Platform
The accurate alignment of cannon barrels in battle tanks is paramount during the engagements where precision can determine the outcomes. In chaotic combat environments, ground vehicles endure intense vibrations from rough terrain, engine operations, and enemy fire, all of which can misalign gun barrels and compromise targeting accuracy. The running engine and transmission further exacerbate these vibrations, posing a threat to mission success and crew safety. Therefore, maintaining precise barrel orientation despite these challenges is crucial for enhancing the lethality and survivability of armored forces. This research proposal aims to develop robust control techniques that can achieve and maintain the desired orientation of gun barrels, effectively counteracting the detrimental effects of battlefield vibrations. The focus will be on designing a nonlinear control law capable of mitigating the impact of environmental factors on the battlefield, ensuring the consistent and accurate direction of the gun barrel. Maharashtra Drone Mission
The Maharashtra Drone Mission (MDM), launched by the Government of Maharashtra, aims to bolster indigenous technology and position Maharashtra as a global leader in drone technology. This statewide initiative integrates academic institutions, research bodies, government agencies, industries, and youth to harness drone technology for the state's development. Headquartered at IIT Bombay, the mission focuses on establishing a fully autonomous drone ecosystem with distributed autonomy. This advancement is crucial for seamless integration into airspace and communities, enhancing efficiency and safety. MDM emphasizes drone system design, secure communication, operation management, and technology development tailored to specific applications across various sectors. It also includes initiatives like droneports, testbeds, skill development, entrepreneurship, policy frameworks, human resource development, and outreach programs. The mission envisions comprehensive development and deployment of drone technologies to propel Maharashtra forward in the global drone industry. Resilience of Vehicular Swarms to Malicious Attacks: A Robust Control Approach
Networked Dynamical Systems (NDS) consist of multiple agents that collaborate through communication to achieve shared objectives. Examples of such systems include transportation networks, power grids, and drone swarms. However, these systems are vulnerable to disruptions like communication failures, agent malfunctions, or external interferences, which can undermine their performance and stability. Therefore, it is crucial to develop resilient strategies that ensure these systems can operate safely and effectively. In addition to traditional robust control methods, this project aims to leverage advanced tools from dissipativity theory, which encompasses concepts like passivity and small gain, to identify suitable storage functions for analysis. The focus will be on addressing robustness and resilience challenges in swarms of diverse vehicles, including both ground and aerial types. The effectiveness of the proposed robustness measures will be validated through experiments conducted on real-world setups. Extension of Swarming Algorithm for Three-Dimensional Scenarios and Validation of its Reliability to Real-world Scenario
This primary objective of this project is to create and showcase UAV flocking algorithm in the three-dimensiaonal scenarios through MATLAB simulations. The project aims to develop algorithm that will simulate coordinated drone behavior by implementing decentralized control and following specific rules and constraints. Our goals include conducting a stability analysis to establish the conditions for consistent flocking and investigating the algorithm's performance under communication challenges like latency and packet loss. In summary, our project aims to deliver a robust flocking algorithm demonstration, stability insights, and an assessment of real-world communication effects on drone coordination. Three-Dimensional Navigation of Autonomous Underwater Vehicle
This project is about 3D navigation of autonomous underwater vehicle (AUV). Autonomous underwater vehicles deployed to assist oil rigs or monitor underwater cabling may differ in the application domain and the environment in which they work. The AUVs are also deployed to track marine life and underwater explorations. Each application of AUV has to address a navigation task that involves moving the AUV from a source location to a target location. The navigation task is typically performed by describing a path between source and intermediate goal locations called waypoints. Depending on the application requirement and AUV design constraints, the waypoint navigation classifies under position or posture stabilization in control terminology. Posture/point stabilization refers to driving the complete state to attain the desired configuration, while position stabilization only concerns stabilizing the robot's position to a target location. Position stabilization is a special case of autonomous navigation, which requires the robot to know either the exact goal coordinates or relative range and bearing measurements towards the goal location to perform the task. In robotics literature, position stabilization is referred to as homing particularly when the onboard sensors do not give direct information on coordinates or relative bearing/range measurements. This project aims to develop a 3D motion planner for the autonomous navigation of underwater vehicles. The output of the planning module would be optimal, feasible reference trajectory to reach the goal point with online obstacle avoidance. Obstacles can be static as well as dynamic in nature. This project also ivolves generation of the required environment map from SONAR data (range, bearing, and track). The validation of propsoed designs will be done using a simulation environment and on an underwater test platform. Capacity Building for Human Resource Development in Unmanned Aircraft System (Drone and related Technology): Guidance, Navigation, and Control Algorithms and Simulations
Unmanned aircraft system (UAS) encompasses unmanned aerial vehicles (UAV), also known as Drone, includes related technologies such as ground control stations, data links, and other support equipment. The technology has the potential for greater reach with better work productivity and relatively lower cost through diverse operational and physical characteristics involving operating range, payload, operational altitude, take-off weight, endurance or flight duration, command & control, etc. The primary objective of the programme is to leverage collaborative activities in human resource development through capacity building in education and training in the area of UAS. The programme is conceived to achieve the following broad objectives: To enhance capacity & capabilities of select institutions in identified work themes on unmanned aircraft systems. To institutionalize a collaborative ecosystem for synergy of capabilities & expertise. To foster the development of competent human resources at various levels, including Post Graduate & Graduate programs, PG Diploma/Certificate programs, Faculty Updation, and Master Trainers in niche areas of UAS. To promote an entrepreneurial mindset and nurture technical talent among the student community through innovative interventions such as Bootcamps and Proof-of Concepts. To nurture technical talent and ideation among the student community through IPR generation, Competitions, Workshops / Conferences, etc. The project has conceptualized to leverage and augment the capacity and capability of academic and related institutions through a unique collaborative framework. The programme is being framed to further strengthen the identified institutes with a mission to create quality human resources which contribute skilled professionals and workforce to the UAS industry. Fault Tolerance Control System for Gas Turbine Engine
Gas turbine engine is a key component for aerial vehicles. The performance of engine plays a vital role in the success of mission. The project aims to design an integrated robust fault diagnosis and isolation (FDI), and fault-tolerant control (FTC) system for gas turbine/jet engine in different fault scenarios. The main objective of fault-tolerant control is to maintain the specified performance of a system in the presence of faults. With the increasing demand on reliability and safety in gas turbine/jet engines, which are prone to components faults and operational abnormalities, it is extremely important to detect and diagnose potential faults and abnormalities as early as possible, and implement fault-tolerant control to minimize performance degradation and avoid dangerous accidents. Advanced FTC schemes are required to ensure efficient and reliable operation of complex technological systems. Fault detection and isolation forms a vital part of any integrated active fault-tolerant system. The development of mathematical models of engine, actuators, and sensors will be required for designing FDD and FTC system design. The model development is proposed to be done using system identification and/or computational techniques. Proof-of-ConceptObject Transportation Using Cooperative Unmanned Aerial Vehicles
In this proof-of-concept, we aim to develop robust guidance and control schemes for a team of unmanned aerial vehicles (UAVs). The proposed strategy will enable the UAVs to transport a load cooperatively from one location to another. Obstacle Avoidance for an UAV Using Collision Cone Approach
In this POC, we aim to design a controller for a quadrotor to ensure collision-free navigation in the presence of obstacles. The primary focus is to utilize the collision cone approach to create an obstacle-free path for the quadrotor. In order to fulfill the assigned task, especially in the indoor environment, a quadrotor might need to travel from one point to the other. While doing so, more often than not, the quadrotor might have to circumvent the stationary and moving obstacles in the environment. Avoiding these obstacles becomes critical for any mission to ensure the safety of the quadrotor and its surrounding environment. The collision cone approach provides an elegant solution to the obstacle avoidance problem. Using this approach, one can obtain the range of heading angles of the quadrotor for an obstacle-free path in the environment. This guidance scheme can be integrated with the vehicle’s control system to adjust its velocity and direction dynamically, ensuring smooth and safe navigation. Convoy monitoring schemes for Various Autonomous Vehicles
A swarm of agents/drones are required to navigate through a 3D space, where the path to be traversed is known only to selected few drones, termed as leaders, while the other drones are called followers. The containment control problem requires the leaders to traverse along a reference trajectory provided to each of them while the followers are required to remain inside the convex hull of the leaders, as illustrated by the representative diagram in the adjoining figure. It is assumed that the leaders can navigate autonomously along a given polynomial trajectory that is pre- specified and the followers only have access to the relative position information between themselves and their neighbours. The leaders, as characterized here, can be assumed to be reference signals and every follower receives state information of the leader, either directly (i.e. through a directed edge in the graph describing the network topology between any of the leaders and itself) or indirectly (i.e. through a directed path between itself and at least one of the leaders). The main contributions of this proposed work may be summarized as follows. Followers’ dynamics should not be required to be of lower order. Despite this, they should be able to converge within the convex hull specified by the leaders' positions, while leaders can follow a polynomial trajectory of any degree. Furthermore, unlike many existing results, the present work should not require the followers to be modeled by identical dynamics. Thus, each follower may be modeled as some second order linear system (some of the followers may even have inherently unstable dynamics). Convoy Monitoring Schemes Consisting of Heterogenous Vehicles
In this POC, we aim to develop a robust control algorithm to monitor the convoy of ground vehicles using a team of unmanned aerial vehicles (UAVs). The UAVs may be equipped with cameras and other necessary sensors for convoy monitoring. The team will be in a particular formation such that the convoy never goes outside the field of view of the UAVs. An illustration of the proposed work is shown in the figure below, where two quadrotors are maintaining equal height with an angle from a ground vehicle doing surveillance over a moving ground vehicle. This concept will be extended for multiple UGVs and UAVs. Completed ProjectsMulti-Vehicular Cooperative Pursuit and Evasion Guidance Strategies: Multi-Agent Perspective
This project investigated multi-vehicle engagements, such as cooperative aircraft defense against multiple adversaries, and propose cooperative pursuit and evasion guidance strategies by harnessing the analysis tools from multi-agent systems, nonlinear systems, and cooperative control. The past decade has witnessed substantial advancements in the area of guidance and control of aerial vehicle. Sophisticated interceptors posed a serious threat to the survival of civilian, as well as, military vehicles. Consequently, significant efforts have been devoted to extending the protection capabilities of the targeted vehicle, which may respond to the incoming threat by performing evasive maneuvers, such as deploying electronic countermeasures, flares, and decoys. While the aforementioned survival tactics employed by the targeted vehicle are passive, an active protection strategy would involve launching a defending interceptor or a team of interceptors, having capabilities comparable to that of the incoming threat, against the adversary. This is a useful requirement, particularly in the scenarios when the targeted vehicle has some constraints on its route, or is carrying a load, such as goods or people, heading for search and rescue, etc. The resulting three-player engagement scenarios are different from the traditional one-to-one engagements, and present a formidable challenge in the design of defense strategy, as the evader-defender team may use different levels of cooperation to neutralize the threats. Therefore, in this multi-vehicle pursuit-evasion scenario, while the targeted vehicle attempts to throw the pursuer off its trail through evasive maneuvers, the team of defenders it deploys tries to pursue the attacker (incoming threat) and neutralize the threat posed by it. The evader cooperated with its defenders to ensure interception of the pursuer by the defender before the pursuer could capture the evader. In this scenario, the evader and the defender were a cooperating team, while the pursuer was the adversary. From the perspective of the defender-evader team (which attempts to ensure not only evasion by the evader but also the capture of the pursuer by the defender), the design of suitable strategies for the adversaries is a challenging problem even from a geometric formulation of the problem. From the point of view of safety applications, safeguarding one’s own aircraft from attacks by pursuing interceptors is important, just as intercepting an enemy’s interceptor is a primary goal. This project, therefore, focused on the former under the unifying umbrella of active aircraft defense, which involves both pursuit (by defenders) and evasion (by targeted aircraft) strategies. Hardware Implementation and Validation of Swarming Algorithm
A swarm of unmanned aerial vehicle (UAV) is a collection of aerial robots working together to achieve a common goal. These systems have immense potential to expand the mission domain of UAV systems by delivering better resilience and adaptability at a lower cost than monolithic systems. With developments in miniature UAVs, the problem of flocking has gained significant interest in both academia and industry, with applications in intelligence, surveillance, and reconnaissance (ISR), electronic warfare (EW), and communication missions. The goal of the project was to design and validate flocking control laws for fixed-wing multiple agents and demonstrate their performance on robustness on appropriate hardware test beds. The final goal was to demonstrate the efficacy of the developed control laws on fixed-wing agents and extensively validate them over various mission profiles. Development of Flight Guidance and Control System of Aerial Vehicles for Path Following
Guidance system acts as the brain of aerospace vehicles, and success of mission relies heavily on the performance of guidance system. In the early days, achieving target interception with zero miss distance has been the primary focus of guidance design for many years. In recent days, optimizing energy usage and achieving terminal constraints are some of the additional constraints, which are of paramount improtance. Terminal constaints such as impact angle increases the warhead effectivenss while the impact time helps in saturating the target defense system and thus increasing the survivability of interceptor. This project focused on designing guidance and control of autonomous aerial vehicles for acheiving terminal as well as in-flight constraints. The terminal constraints include target interception at a particular impact time or interception from a pre-sepcified orientation, while the in-flight constraints accounts for field-of-view (FOV) constraints which helped to keep the target in view during engagement. This project aimed to design both two-loop as well as integrated design of guidance and control to achieve the objectives. Guidance design must account for the variations in vehicle dynamics subject to the thrust and aerodynamic parameter variations. Owing to the nature of engagement duration, the developed guidance strategies respected the finite time convergence of controlled variables. Cooperative Nonlinear Guidance Strategies for Simultaneous Interception with Finite-Time Convergence
This project was about the design of cooperative guidance strategy for several vehicles to achieve simultaneous interception of the target. Modern targets are becoming technologically advanced which imposes several challenges for vehicles to complete their mission. Simultaneous interception by weaker adversaries against a stronger one are effective. This project aimed to design cooperative salvo guidance strategy using finite and fixed-time consensus in coordination variables. This project will also attempt to develop cooperative salvo guidance, against various targets, at a time either pre-specified or decided online. Guidance philosophy would be based merging the concepts of consensus in multi-agent systems and nonlinear control techniques to cater for the communication capability of vehicles, along with the robustness against the erroneous measurements/estimation of data and uncertainties.
Completed Proof-of-ConceptsNavigation of UAV Through Arbitrary Openings
Our primary objective in this POC was to design and implement a robust algorithm for an autonomous unmanned aerial vehicle (UAV). The algorithm's fundamental goal was to enable the UAV to navigate through an arbitrarily shaped opening in a wall. To accomplish this, the algorithm should fulfill the following objectives: 1) The UAV should be capable of detecting the edges of an arbitrarily shaped opening in a wall. The algorithm should then analyze the detected edges and connect the dots to form the opening contour accurately. 2) After identifying the shape of the opening, the algorithm should incorporate a feasibility check to determine whether the UAV can pass through the opening while maintaining an optimal safe distance from its edges. This additional step ensures that the UAV can assess the dimensions of the opening and make informed decisions regarding its navigation, prioritizing safety and avoiding potential collisions with the edges of the opening. 3) By incorporating these techniques, the UAV can navigate the identified opening precisely and autonomously. Robust Guidance Scheme for Fixed-Wing UAV
In this POC, we aimed to develop a robust guidance system for autonomous path-following of fixed-wing unmanned aerial vehicles (FWUAVs) with the following capabilities. Autonomous path-following, that is, the vehicle can follow a predefined path in the three-dimensional space without any human interventions. The FWUAV can converge to its desired path and follows the same for all future times in the absence of path curvature information, thereby made the proposed design effective for any generic path. Remain effective in the presence of wind, as wind is inherent in the environment. Therefore, the proposed guidance systems had capabilities to handle the wind. Trajectory Tracking of Multi-copters in Constrained Space
In this POC, we aimed to develop a robust control system for autonomous trajectory tracking of a multi-copter with the following objectives. The vehicle could converge to its desired trajectory at an exact time instant in the presence of a partially known model and exogenous disturbances, which may emulate the effect of wind and other environmental imperfections. Vehicle ensured precise tracking in the presence of yaw, acceleration, and deacceleration constraints. The proposed control design steered the multi-copter to track the desired trajectory even if a spatial constraint is imposed on the movement. |