DRΛGOON Search and Rescue Robot

This project consists of the end-to-end design, engineering, validation and testing of a robotic system taking place over 3 semesters of my degree. Beyond simply bringing a robot to fruition, this project particularly focuses on the systems engineering and project management processes required to elicit a high quality product in minimal time. From requirements engineering to final system validation, the output of this effort is far more than just the single robot.

In partnership with Draper Laboratory, my team is developing a ROS-based search and rescue ground robot to conduct autonomous interior mapping and survivor localization within a disaster environment. I am the technical lead for human detection under adverse visual conditions, which includes minimal lighting and obstacle occlusion, in both low resolution thermal and RGB imagery. The constantly evolving primary website for this project can be found here, and the preliminary Critical Design Review Report found below.

Contributions and Responsibilities

• [Lead] Training, testing, acceleration and integration of deep learning for human detection in RGB and IR imagery

• [Lead] Integration, calibration and testing of IR camera

• [Lead] Multi-hypothesis Kalman filter integration for human localization

• [Lead] Off-board operational visualization development and design

• [Lead] 3D modelling and rendering, graphic design and visual storytelling

• [Assist] Structural design and fabrication for custom camera, LIDAR and AGX mounts

• [Assist] Software design for human detection and localization pipeline

• [Assist] Full-system management, design, engineering, testing, integration and demonstration