Dilli Bhaskar

Student Researcher

Email / GitHub / Linkedin / CV

About Me

I’m a Final Year undergrad in Mechanical Engineering at TCE. Currently I'm interning at Control and Robotics Laboratory at IIT Pallakad supervised by Dr. Shaikshavali Chitraganti. Previously I have worked at Artificial Intelligence lab supervised by Dr. M.Suguna. where, I built a presice object detectors on Themal images

I am also interested in the learning dynamics of Deep Learning agents, using mathematical tools from optimization, dynamical systems and learning theory to understand what RL agents learn (and fail to learn).

Research Interests

  • Robotics
  • Deep Learning
  • Model-free Reinforcement Learning
  • Deep Reinforcement Learning

Projects

Early Thermal Forest Fire Detection using UAV and Saliency Maps

• Fire detector was designed using YOLO for Thermal images.
• Saliency Maps of the Thermal images was obtained using Deep Saliency Network (BAS-Net) and fused with thermal images. Using Yolo-v7 object detection is performed on the regenerated fused thermal images.

|Paper| Details|

Street Litter Detection

• Built waste and person detection using Yolo-v5 model to identify the street litter based on euclidean approach.
• Trained on Pennfudan & TACO Dataset and achieved a F1 score of 89%.

|Video| Details|

LQR contolled Self Balancing Bike

• Designed and Modelled self balancing bike in Solidworks and V-rep for autonomous delivery of items.
• LQR controller was used to self balance and IR sensors were used for line following.

|Video| Details|

AI-assisted Waste Recycling System

• Smart India Hackathon 2022 Award Winning Project.
• Built three DL models, for identification of waste on the road, to classify it into seven categories for recycling and disposal, to classify the class plastic into five sub categories based on its shape.
• Implementation is in progress in collaboration with ICCE.

Details

Design Analysis of Detroit 8V-71 Marine Diesel Engine Piston

• Re-designed Detroit 8V-71 Engine using Convex and Concave Profiles for Piston head.
• Using Ansys Workbench static stress analysis was performed and observed that Concave Head profile shows better result than Flat Head profile and Dome Head profile in terms of stresses and total deformations.

Details

IOT integration on Shaker Machine

• Developed low cost non-contact AC sensor for check the status of components of shaker machine available in Product reliability lab, TSS CAR , TCE.
• Sensor data are processed and deployed in Arduino IOT Cloud platform and Google Drive for long time usage

Video

Aerodynamic Drag Study of Time-Trial Cycling Helmets

• Made flow analysis of cycle helmet with tail flap of 0, 3, 9 degrees to the horizontal using Ansys Fluent.
• Conclusion made was tail flap position of 6 & 9 degree to horizontal, produces a considerably low drag coefficient.

Details

Design and Fabrication of Miniaturized Aircraft

• Fabricated rubber band powered aircraft using Balsa wood & Paper.
• Dihedral angle : 20degrees
• Time of Flight: 2mins 10s

View Project