Robotics 2018 SURE and SROP Research Projects

Robotics Project #1 (for 2 students): Development and Testing of Robust Flight Control Systems and Experimental Platforms
Project Mentor: Peter Gaskell, pgaskell@umich.edu
Prerequisites: Significant programming experience. Experience with unmanned flight systems (i.e. multi-rotors or RC planes)
Project Description: We wish to design implement several unmanned aerial platforms for experimentation with sensors, control systems, navigation and planning. The project will involve significant software development in C/C++ and the design and assembly of one or more unmanned aerial vehicles. The project will also involve flight testing of the system so RC flying experience is a plus.

Robotics Project #2: Rehab Robotics: A Powered Exoskeleton for Improving Ankle-Foot Orthosis Prescription
Project Mentor: Kira Barton, bartonkl@umich.edu
Prerequisites: Experience with human research and/or motion data analysis in programs such as Visual 3D is a plus.
Project Description: An ankle-foot orthosis (AFO) is a brace worn by a variety of patients with abnormal walking patterns, including multiple sclerosis, stroke, and cerebral palsy patients. AFOs support the movement of the ankle joint and help correct deformities. However, there are many different AFO designs to choose from and the current fitting process at the clinic lacks quantitative patient performance metrics. Our lab has built an exoskeleton system that can mimic the behavior of different AFOs and allow patients to ‘try on’ different designs by wearing a single powered exoskeleton boot. The system will use real-time patient data from various sensors to help clinicians decide on an optimal AFO design to improve patient outcomes. The SURE student will help run experiments to validate this system with healthy and patient subject populations, as well as help analyze the collected data.

Robotics Project #3 : Active Perception for Manipulating Piles of Objects
Project Mentor: Dmitry Berenson, dmitryb@umich.edu
Prerequisites: Significant programming experience in Python and C++. Experience with computer vision and/or robot manipulation is a plus.
Project Description: As part of a joint effort with Prof. Jason Corso’s group, we are investigating how a robot can take actions to gain more information about the environment. Specifically, when a target object is occluded in a pile, which other object should we remove to get a better view of the target object? Another question is how to remove the occluding object; i.e. picking it up, pushing it aside, etc. This project will allow students to integrate methods in both manipulation and perception into a unified framework.