Project: #178 Aerial Contact Sensing for Improved Inspection of Transportation Infrastructure Progress Report - Reporting Period Ending: March 30, 2019 Principal Investigator: Sebastian Scherer Status: Active Start Date: July 1, 2018 End Date: June 30, 2019 Research Type: Applied Grant Type: Research Grant Program: FAST Act - Mobility National (2016 - 2022) Grant Cycle: 2018 Mobility21 UTC Progress Report (Last Updated: April 1, 2019, 8:03 a.m.) % Project Completed to Date: 75 % Grant Award Expended: 75 % Match Expended & Document: 75 USDOT Requirements Accomplishments Construction of an Aerial Platform We have chosen a novel design for a MAV that is capable of full 6 degree of freedom motion. This means the MAV doesn't need to change its orientation to regulate its position, as is the standard. It makes use of a rotors that are tilted at a fixed angle, so that they can generate thrust in the X and Y directions of the UAV. We have built the platform and tested it in position control mode. This involved a state estimator based on GPS and IMU was used as feedback for the control. We were able to verify that the MAV can be maneuvered without needing to ever change its orientation. A robotic arm prototype was developed and built. It is based on the uARM design (https://www.ufactory.cc/#/). This design was chosen because it concentrates the motors on the arm base, which decreases the inertia on the arm, and because its parallel mechanism guarantees the end effector pose is always at a fixed relative orientation with regards to the arm base. Simulation Development A simulation has been developed based on ROS and Gazebo that is able to recreate both the peculiar aircraft design and the robotic arm. The simulation has been show to be a good approximate for the vehicle flight dynamics, and allows complete testing of all the perception and control stack, which have ROS-based components but also uses the PX4 development framework. Initial Control Algorithms A strategy for MAV control in close proximity to structures has been implemented and tested in simulation. It is uses sensor feedback from a camera and laser mounted on the MAV. It dispenses any GPS information, since its quality is not good enough in these scenarios. The next steps are to validate this strategy on the aerial platform. Impacts N/A Other None Outcomes New Partners None Issues None