My capstone project at Carnegie Mellon University, supervised by Dr. Oliver Kroemer, is to develop an autonomous food assembly kiosk that can prepare and serve a variety of dishes to customers. The project involves developing a vision system to detect and track ingredients, a robotic arm to pick and place ingredients, and a user interface to take orders.
I am working on a project to develop an autonomous drifting algorithm for an F1-Tenth RC car using Deep Reinforcement Learning. The project involves developing a simulation environment, training a reinforcement learning agent, and deploying the agent on the physical car.
I collaborated in a team of five to build a ROS2 based autonomous car that can navigate a track using LIDAR input.
I implemented a Gap Detection and Following algorithm that determines speed and steering angle. We demonstrated our car at the Safety21 Symposium at Carnegie Mellon University. As a group, we demonstrated our work to former secretary of transportation Pete Buttigieg.
I independently developed vision algorithms for a dual 6-axis DENSO VSA manipulator system to handle garment pieces as part of a clothes assembly line at The Center For Transformative Garment Production, Hong Kong
I contributed to the research, design, implementation, testing and deployment of closed-loop PID control algorithms for Level-1 ADAS features such as position holding, heading holding and holonomic motion on electric outboard motors meant for light surface vessels at ePropulsion Technology Limited, Hong Kong
At Avalon SteriTech Limited, I implemented Simultaneous Localisation and Mapping (SLAM), Waypoint Navigation, Collision Avoidance, Human Detection, Remote Operation and Autonomous Workflows on a quad drive wheelbase extended with a disinfectant spraying module for clients such as the Hong Kong Mass Transit Railway (MTR), Cathay Pacific and HAECO.
Performing computer vision tasks such as object detection, image segmentation and object tracking using neural networks is a skill I have honed over three different projects at Kodifly Limited, Hong Kong.
Dog Detector for Mass Transit Railway corporation (Hong Kong):
Real-time detection and tracking of dogs on surveillance feed from MTR stations.
Counting and Identifying the detected dogs
Medical instrument counting for Hong Kong Science and Technology Parks:
Identify and count different types of medical instruments placed in a tray to speed up
surgery preparation.
Restaurant Activity monitoring for undisclosed client in Hong Kong:
Identify and track each unique employee’s activities
Ensure employees are practicing safety procedures (during COVID-19)
Detect running taps, open refrigerators, unnecessary operation of lights/fans
Non-Disclosure Agreement prohibits sharing of media
At RaSpect Limited, the product ‘Inspectia’ is an inspection and visualization software system that analyzes visual and sensory data to create a safety and health report on building facades and interior structures such as elevator shafts. As an AI Engineering intern, I developed a depth perception application and integrated it with the DJI Manifold 2 onboard computer and the ROS API of the Matrice 300 RTK drone.
Non-Disclosure Agreement prohibits sharing of media
At ClearBot Limited, I developed a ROS2 driver node to interface with electric outboard motors, sending throttle and steering commands and retreiving feedback such as power, voltage, current, temperature and RPM.
Non-Disclosure Agreement prohibits sharing of details
At Hanson Robotics, I experimented with Artificial Intelligence to extend Sophia the robot’s capabilities, reduce interaction time and enhance user experience.
Non-Disclosure Agreement prohibits sharing of details
I volunteered at MakerBay Hong Kong, and in collaboration with Hong Kong Open Ocean Youth, developed two prototypes of unmanned surface vehicles to collect images of coral reefs along with GPS location stamps
I identified that recycling rates in Hong Kong were alarmingly low due to lack of efficient trash sorting. To address this problem, I created a prototype for a trash sorter that uses a Yolo V4 tiny network running on Raspberry Pi that processed visual feed of the trash and classified it into paper, plastic, metal and glass categories
The Vayu project was started long before I joined it and its aim was to create the world’s fastest robot fish. When I joined in 2020, it was with the founding of BREED Robotics, a student society at HKU to nurture the creation of bio-inspired robots. As co-founder and technical leader of the electronics division, I oversaw the design and implementation of the circuit of the final version of the fish.
Hi there! I am Abhinandan, or Abhi, as my non-Indian friends call me. I am an Engineer with a passion for Robotics and Artificial Intelligence.
Having worked at several companies on various Robotics, Computer Vision, Deep Learning, Embedded Software and IoT at different stages of product development, I am currently sharpening and updating my skill set via a masters program in robotics at Carnegie Mellon University.Having contributed to technology that has transformed industries, benefited societies and helped conserve the environment, I am eager to continue building and shaping the future keeping in mind sustainability, efficiency and eco-friendliness.