Credit to Misha Krutiy for producing the video.

autonomappr.com

Category : Engineering / Capstone
Completion : September 2017 - March 2018
Role : Test Lead
Tools : MATLAB | SolidWorks | Python | ROS | DroneKit

Autonomappr is a system that enables autonomous vehicles on construction sites. It provides up-to-date collision map data as well as geographic info about temporary roads and points of interest. Construction managers are able to enter road and point of interest data into Autonomappr's website as things change on the side through day to day activities.

As the test lead I was responsible for analysing test data collected from various sensors, as well as the image data to determine system's performances. Before the drone was ready I got a simulation of the drone working in ROS and DroneKit API. The simulation allowed the team to verify the automated flight path generation as well as path safety for the drone. Pre drone construction I spent a lot of time on research for potential semantic segmentation of the data that would be collected by the drone. Once the drone was ready, I spent majority of the time analyzing the trends in the collected data, and making deductions about the effect of parameters on the designed system cost, time, efficiency, anti vibration mount and other factors.

Download Design Docs Download Symposium Poster Download Final Presentation
← Previous Project Next Project →
pshering@edu.uwaterloo.ca