Robotics Engineer
Published:
Wastefull Insights, a deep-tech startup headquartered in Vadodara, is pioneering AI-driven solutions for effective waste segregation. The company introduced "Achhe Dinn", India's first e-commerce platform promoting recycled, upcycled, and sustainable products. Focused on source segregation, circular economy practices, and automating dry waste segregation, Wastefull Insights is driving change across the nation.
Internship Experience
My journey with Wastefull Insights began as a Robotics Engineering Intern, where I took on the following tasks:
- Studied the existing code base to gain a deep understanding of their systems.
- Developed a scheduling-based optimization algorithm for a pick-and-place segregation robot.
- Explored ROS Serial and Tiva boards, contributing to a marker-based conveyor belt velocity estimation algorithm.
Key Technical Contributions:
- Scheduling Algorithm:
- Designed a system combining a Kalman Tracker with a YOLO object detector for optimized scheduling.
- Velocity Estimation:
- Developed a solution leveraging ArUCo markers for conveyor belt velocity estimation.
After successfully completing my internship, I transitioned into a full-time Robotics Engineer role.
Full-Time Role: Robotics Engineer
As a Robotics Engineer, I undertook projects that defined the next phase of my professional journey:
- Code Base Migration:
- Migrated the entire robotics code base from ROS1 (Python) to C++.
- Designed a modular system architecture enabling multiple robots to operate simultaneously on a single conveyor belt, improving efficiency and scalability.
- Independent Systems Development:
- Addressed limitations in ROS by disintegrating specific modules into independent C++ programs.
- Developed a CI/CD pipeline to ensure robust system portability and seamless updates. This was crucial for maintaining the performace of robots at company premise and deployment facilities.
- Vision Pipeline Optimization:
- Conducted initial experiments on Jetson Nano and transitioned deployments to Jetson NX for enhanced performance.
- Transitioned the vision pipeline from a frame-by-frame approach to a streaming process using the Nvidia DeepStream SDK, significantly enhancing processing efficiency.
Reflections and Gratitude
My tenure at Wastefull Insights was a transformative experience, providing me with hands-on exposure to cutting-edge technologies and interdisciplinary collaboration. I owe my deepest gratitude to Rishabh Shah and Manali Agarwal for their unwavering mentorship and support, which were instrumental in my decision to pursue graduate studies.
A big shoutout to my exceptional colleagues – Mayank Bansal, Ashwin Chakravartula, Sachin Purani, Tejas Phutane, Henish Viradiya, Nishith Mehta, and Hrithik Dhongadi. Their camaraderie and teamwork turned challenges into opportunities and made every achievement memorable.