Simulation of Walking with a Robotic Exoskeleton

Goal: Analyze effects of HAL on human gait using musculoskeletal modeling and simulation

  • Create a model of Human wearing HAL
  • Use motion capture, EMG and forceplate data to simulate motion of human walking on treadmill wearing HAL to study kinematic parameters
  • Study muscle activation patterns
  • Use Inverse Dynamics tools to study joint forces and torques

Building an Instrumented Glove

Goal: Obtain kinematic data including position, velocity, acceleration of a subject performing a grasping task

The glove is equipped with MEMS based accelerometers at the finger tips and infrared markers to give complete information about the grasping task.

 

Optimizing Performance of Reciprocating-Fork Hopper Feeder

Goal: Use factorial design to systematically evaluate performance of Reciprocating-Fork Hopper Feeder and suggest optimal operating parameters.

feeder_1

 

 

Impact Attenuator Design for FSAE Vehicle

Goal: Design an impact attenuator to be mounted on the front bulkhead of the vehicle.

 

fsae_1

I was in-charge of the Vehicle Safety design subgroup in the first season, and my primary responsibilities were:

  • Conception and validation of an aluminum honeycomb-based Impact Attenuator for the vehicle as per competition specifications
  • Acquisition of honeycomb sheets, and fabrication of the attenuator
  • Preparing and presenting the reports on vehicle safety during static test events

Additionally, as it was the first season of the team, I also actively participated in

  • Marketing and publicity – we secured 90% sponsorship for the purchase of aluminum honeycomb sheets
  • Design of team brochure

Accolades for the team in Formula SAE India 2009 Design Challenge

  • First successful vehicle roll-out from NSIT
  • Widely appreciated for the design of impact attenuator at the event
  • Ranked 4 th in Design Evaluation, 4 th in Design Presentation and 6 th overall (out of 28 teams).

fsae_2

 

Intelligent Automotive Suspension System

Goal: Investigate the performance of an active suspension system with PID and Fuzzy Logic Controllers

Design of a social robot using LEGO Mindstorms

Goal: Design a socially interactive robot based on cognitive neuroscience principles

 

Using Neural Networks for Predicting Closing Values of NIFTY index

Goal: Use Artificial Neural Network (ANN) to model closing values of the National Stock Exchange Index of Indian (NIFTY).

A 2-layer neural network with backpropagation is trained on historical data from 2000-2014,  and  its performance is evaluated for varying number of neurons in the hidden layer of the network.

 

 

Optimization of Dry Etching of GaN using Taguchi Method

Goal: Optimize low pressure dry etching of a- plane and c-plane GaN using Taguchi arrays.

vsrp_1

 

vsrp_2vsrp_3