Robotics and automation systems are integral for developments in manufacturing, healthcare, safety monitoring, transport and many other fields. Our faculty is involved in innovative research in robot locomotion and coordination, soft robots,rehabilitation, compliant mechanisms, metamaterials, multifunctional materials, haptics, acoustics, disaster management, structural health monitoring and machinery diagnostics. Their research builds on advances in machine learning, sensors and materials. They collaborate with industry partners to execute cutting-edge projects and solve real-world industrial and societal problems.

● Robot locomotion and coordination                                    
● Soft and compliant manipulation systems                                    
● Smart materials and structures                                    
● Haptics and tactile interactions                                    
● Vibrations, ultrasonics, acoustics and hearing research                                 

Robotic locomotion and coordination                                    
Anirban Guha, Shiva Gopalakrishnan, Abhishek Gupta, Vivek Sangwan    

Locomotion:     

Wheeled Autonomous robots                                    
Prof Anirban Guha’s lab studies and designs locomotion strategies for autonomous robots and remotely operated vehicles (rovers). They design and build robots and rovers along with the algorithms to efficiently maneuver them. The team has analysed the suitability of a commonly used rover locomotion mechanism, called “rocker bogie”, for step climbing. Their findings provide valuable insights on how link lengths and wheel radius can be selected according to the step height. Guha’s team has previously developed a remotely operated vehicle for the Indian Army.           

The team has developed an operating algorithm for an onion harvesting agricultural robot that considers the battery status while planning the path of travel. They demonstrated the safe return of the autonomous robot in scenarios where the battery depletes normally as well as when there is fault in the battery while harvesting. A planning strategy that incorporates battery management and suitable path planning methods is necessary to ensure safety and energy efficiency during operation.

Modelling and controlling the motion of legged robots;                                   
Prof Vivek Sangwan’s work is focused on studying motions such as walking and acrobatic moves in two-legged robots. Sangwan’s team has modelled and analysed the walking trajectories for biped robots. Their work has helped in creating a library of feasible trajectories covering a range of speeds, control methods and motion constraints, from which the most optimal trajectory can be chosen quickly, depending on the conditions and the time of operation, without additional computation. The team has also explored complex acrobatic maneuvers, developing two actuation strategies for a simple two-link robot to perform handsprings. The two-legged robot work is supported by DST-SERB.

Rehabilitation robots                                   
Robots are used to enable and support movement during rehabilitation. Prof.Abhishek Gupta’s team designs wearable robotics devices to aid people, especially below-knee amputees, in walking. A prosthetic device developed in their lab uses series elastic springs, which help in achieving a more natural motion, akin to the movements of the human ankle and knee, compared to existing prosthetics. Moreover, their design adjusts the ankle and knee angles in real-time, enabling a smooth walk.

Multi-legged walking robots                                  
Prof Seshu Pasumarthy’s lab works on developing electro-mechanical systems for walking robots. The team designs walking robots that can change their dimensions to fit narrow passages or crawl through pipes. The team has developed a six-legged walking machine with a robotic arm which can clear objects, pick up samples and take videos in an environment strewn with obstacles. The machine has six legs which it can gather together to go through a narrow passage. The machine can also decrease its height when there’s a low passage.

Robot Coordination                                  
In situations such as in manufacturing and remote exploration, many robots may need to do a task together. We need to ensure that multiple automated elements can work and coordinate well. Facilitating effective coordination between the different elements is a key research focus of Prof Anirban Guha’s lab.          

Prof Guha’s team develops task allocation algorithms to optimise the utilisation of the different robotic manipulators according to their individual capabilities and limitations, based on parameters such as grip strength and stability. Their methods involve collecting data from each robot in the group in real-time to assess their task capability and make decisions on which robots to include or exclude from the task, ensuring effective collaboration in the group. They tested their new methods on robots fabricated from scratch in the laboratory. The key to their success was the algorithm which was much faster than conventional algorithms, while also being computationally less intensive. Going forward, his group plans to use artificial intelligence methods for automation tasks for various industries.

Soft and compliant manipulation systems                                 
R. Ganesh, Abhishek, Prasanna Gandhi

Compliant manipulation systems                                 
Compliant mechanisms are flexible structures, usually consisting of a single piece,that transmit or transform force and motion using elastic deformation. They require low power and no lubrication, are noise free and frictionless. Examples of compliant mechanisms are all around us; in consumer products like tweezers, single-piece bottle lids that snap open or close, push-press buttons to micro- and nano-scale components in automotive, biomedical and various other applications. These types of devices can be used in the place of rigid mechanisms to reduce wear and tear and enhance durability and accuracy.         

Prof Prasanna Gandhi’s lab is involved in developing linear and rotary motion control platforms using compliant mechanisms for various applications like precision positioning, micromilling and micro 3D printing. The team has developed a rotary compliant mechanism for torque sensing in industrial applications that can achieve accurate noise-free measurements without strain gauges. They have successfully used the mechanism to sense output torque of harmonic drives, which are flexible gear transmission systems. They have developed a compliant mechanism based 3-D printing machine to print micro and nano sized components including microneedles for painless drug delivery.

Soft Robots                                
Prof. R. Ganesh and his team in the Dynamics of Flexible Structures lab work on multimodal locomotion for soft robots. Currently available soft robots predominantly mimic unimodal locomotion, i.e. peristaltic rectilinear locomotion or bending-based slithering. The team’s methods combine structural flexibility, and the effect of dynamic excitation to achieve deformation-driven locomotion in soft robots.

Leech inspired soft robot         
Taking inspiration from nature, Prof Gandhi’s team has developed a robot that mimics locomotion of leeches. One end of the robot holds on to a surface with a magnetic or vacuum hold. Highly flexible links connecting the two ends help the non-stationary end to move freely. Leech-inspired robots can find potential applications in window pane cleaning in an energy efficient way.

Smart materials and structures                             
R. Ganesh (Metamaterials), Sripriya Ramamoorthy (Multifunctional Materials), Seshu Pasumarthy (Tensegrity structures)                             
 

Metamaterials                             
Prof R Ganesh’s team at the Dynamics of Flexible Structures laboratory leverages flexible, nonlinear mechanical structures to develop novel multifunctional metamaterials. Metamaterials differ from naturally-occurring materials due to the fact that their mechanical properties also depend on the geometric structure of the material. For example, Auxetic metamaterials display negative Poisson's ratio, which implies that when a tensile load is applied, the material expands instead of contracting in the direction perpendicular to the applied force. The DoFS lab uses analytical, computational, and experimental techniques to develop metamaterials for vibration absorption, as well as for modern day innovations such as wearable electronics and soft robotics. Currently, the team is exploring the dynamics of Kirigami metamaterials, which have a geometric structure inspired by the cuts and folds practised in Kirigami paper craft, to improve existing materials for use in engineering applications.

Multifunctional Materials                           
Sripriya Ramamoorthy’s team designs new types of structures for porous materials that combine multiple functions such as acoustic, mechanical, and thermal functions. For instance, in consumer products like laptops, fans used to dissipate the heat generated tend to be noisy. But space constraints make it difficult to add components to absorb that noise. Integrating both noise and heat control into the same material can keep the laptop small while improving the user experience. Hence, these multifunctional materials are especially useful in applications that have space and weight constraints, including automotive, aviation, and data center thermal and acoustics management, since multiple functionalities can be built into the same material.        

Currently, the team is working on designing a new class of heat sinks using periodic porous materials. These materials have high mechanical strength and can dissipate heat and absorb noise. These materials can then be used for industrial applications that need both noise control as well as heat dissipation. The lab is also developing new techniques for integrating active feedback control with the porous materials to increase the absorption of low-frequency noise in foams. In one recently proposed control strategy, they demonstrated about a 35% increase in the noise absorption coefficient of the foam over a broad frequency range.

Tensegrity structures                           
Tensegrity models are structures where compression from the rigid elements like bars and struts and tension in the flexible elements like cables and wires work together to ensure a stable configuration. Seshu Pasumarthy’s lab studies tensegrity structures that are stable, light and reconfigurable. These structures have applications in robotics.

Haptics and tactile interactions;                         
Abhishek Gupta, Vivek Sangwan                        
 

Simulation of virtual environment through haptics                        
Prof Abhishek Gupta’s work on haptics involves simulating various virtual environments through haptics, the sense of touch. Haptic devices can be programmed to respond with a force output to convey a range of hardness and stiffness like touching a hard wall or the deformation of some object when a user interacts with it. Gupta’s lab develops computational methods to accurately render different hardness and stiffness and investigates how these sensations are perceived by the user in different scenarios to ensure realistic simulations. The team’s recent work demonstrated the effectiveness of their proposed modifications in simulating dynamic interactions through haptics on a Phantom haptic device.

Tactile interaction with drones                        
In some disaster management situations, we might need a drone equipped with a hammer like pendulum that breaks into a building through glass windows and hits an emergency switch precisely. This scenario requires the drone to accurately maneuver itself as well as the tool it carries. Vivek Sangwan’s lab works on developing control strategies for tactile interactions in scenarios that require complex maneuvers like the above. The control strategies are useful when drones are used to perform dangerous tasks like painting walls, cleaning windows of tall structures, power cable inspections and maintenance of bridges.

Vibrations, ultrasonics, acoustics and hearing research                       
Nitesh P. Yelve, Sripriya Ramamoorthy                       
 

Structural and machinery diagnostics                       
Industrial machinery shows changes in their vibration signature when they experience faults such as misalignment in components, loose foundation, unbalance, and damage in components such as bearings, gear teeth or pump impeller. Some of these faults also generate specific frequencies. Prof Nitesh Yelve’s Wave and Vibration Engineering Lab (WaVE Lab), leverages the patterns and frequencies of these vibrations to diagnose faults in industrial machinery. It is important to catch these faults early to avoid compromised safety, machinery failure, and subsequent downtime which may lead to financial losses.        

Traditional fault detection methods struggle when there are multiple faults involved and factors like rotor unbalance, misalignment, fluctuating load, or power supply variations cause nonlinear and nonstationary conditions. WaVE lab develops machine learning, analytical and numerical methods to identify and diagnose faults in such situations. They use mode decomposition techniques to separate complex vibration signals into simpler components and analyze them individually, even when multiple faults are present. Machine learning methods are used to automate the analysis and detect anomalies in the vibration pattern. Machine learning models are also trained on the collected data, enabling them to predict faults before they reach their critical limits. The WaVE lab also uses machine learning to develop digital twins that help diagnose multiple faults in rotating machinery.        

The WaVE lab is equipped with a state-of-the-art industrial fault simulator and data acquisition system to test faults such as inner and outer race bearing faults, horizontal and vertical shaft misalignment, cocked bearings, looseness, gearbox faults, and unbalance in rotating machinery. They have collaborated with various industry partners, including Tata Power and Ingenero delivering high-quality work.        

Prof Yelve’s WaVE lab also does structural health monitoring of mechanical and aerospace structures such as plates, pipes, and composite laminates. The modal parameters of a structure are different for normal and damaged conditions, and thus damage can be detected by evaluating these parameters. The lab is equipped with an impact hammer system and a laser Doppler vibrometer that can be used for vibration-based evaluation of modal parameters of structures. Other methods the lab uses to detect damage in structures are ultrasonic guided wave-based linear and nonlinear methods. By observing the response time and frequency data, the researchers can identify the location and severity of damages. Ultrasonic waves can be guided through wide surface areas, hence the methods can be used to test large areas too. The methods are non-destructive and can detect surface and internal damages even in areas without direct access.

Diagnosis of defects in human hearing                     
Sripriya Ramamoorthy’s lab uses computational and experimental methods to develop a better understanding of human hearing, improved auditory diagnosis and prosthetic devices. The lab uses computational modeling methods to explore sound processing in the human ear such as predicting the quiet sounds called otoacoustic emissions from the cochlea. In connection to this, the researchers are also developing non-invasive auditory methods to measure such emissions.