The EPIC (Exoskeleton and Prosthesis Intelligent Controls) Lab research areas include automation & mechatronics and bioengineering with a focus on the control of powered robotic prostheses and exoskeletons to assist human movement. We implement biological signal processing, intent recognition, and control systems based on EMG and mechanical sensors to improve human-machine capabilities. Our primary goal is to use robotic augmentation technology to restore human movement to individuals with mobility disability.
Issues Involved or Addressed
Lab research is divided into multidisciplinary research subteams of graduate and undergraduate students. Each team is led by at least one graduate student mentor who directs the team to assist in their primary research goals. To best ensure meaningful contributions, each VIP member is selected for a specific subteam. As an applicant, we ask that you first review the subteams (listed below) and reach out to the contacts corresponding to your project(s) of interest. Successfully contacting the graduate student mentors before applying to VIP greatly helps us in determining who to accept! When you submit your application, please indicate the subteam of interest in the comment box.
AUGMENTATION OF PERCEPTION AND ACCELERATED MOVEMENT
Description: This team will primarily focus on developing the testing platform for a hip exoskeleton to explore if this system can assist a user to evade moving threats and running resulting experiments. Once developed the system will work with human subjects to collect real-world data. The team will use an array of sensors for motion capture, muscle signals, IMUs, etc. The project’s long-term aim is to have a plug and play system which can place the user in a virtual environment and throw virtual objects at them for them to react to. We will provide visual, audio, and tactile sensory to elicit human movement. Major short-term goals which undergraduate students can tackle include assembly, testing, and benchtop validation of the hip exoskeleton; development of the virtual environment and other needed components for experimentation; and system integration for making the system untethered from the lab so we can test in any environment. We are seeking students who have interest and/or experience in mechatronic design, programming, and controls. Some of the majors we will be recruiting for are ECE, CS, BME, and ME (If you are ME with a minor in CS this project is a good fit for you). Contact: Aakash Bajpai (firstname.lastname@example.org)
BIOMECHANICAL RISK FACTORS OF LOW BACK PAIN
Description: This team will be dedicated to specializing in relevant biomechanical experiments and analyses in the lab. Undergraduate members of the team will participate in the collection and analysis of motion capture and electromyography (EMG) data. These skills will be used to support a Rubicon sponsored biomechanical study to determine worker injury risks and effective exosuit assistance using the musculoskeletal modeling software, OpenSim. Outside of assisting during human-subject experiments, VIP members can expect to work on projects related to adding wearable robots into simulation, processing biomechanical data, and exploring machine learning techniques. Those interested in human biomechanics, musculoskeletal modeling, and experimental data analysis are encouraged to apply. Contact: Dean Molinaro (email@example.com)
BIOMECHANICS AND WEARABLE SENSING PROJECT
Description: This team focuses on instrumentation with wearable sensors, including IMUs, goniometers, pressure sensors and a novel epidermal flexible EMG sensor. We are interested in analyzing the information carried by these sensors to develop intent recognition algorithms and gait state estimation using machine learning techniques. We performed a full data collection system including motion capture and force plates in our terrain park that includes ramps, stairs and ground level walking. This will allow us to study the biomechanics of ambulation for different conditions and get a better background for the development of controllers for our assistive devices. The focus in the spring will be on flexible sensing experiments and heavy data processing/machine learning/biomechanics processing of data collected on the terrain park. We are open to students from any discipline. Contact: Jonathan Camargo Leyva (firstname.lastname@example.org)
HUMAN-IN-THE-LOOP HIP EXOSKELETON CONTROLS
Description: This team works to understand the link between wearable robotics and the human biomechanical/metabolic system during walking. Using a hip exoskeleton actuated by powerful off-board motors, we are able to test a variety of controllers (including proportional myoelectric, impedance, neuromuscular model based, hybrids, etc.) for their effects on muscles, tendons, and the energetic cost of walking. We understand these effects by collecting and analyzing motion capture data (kinematics and kinetics of motion), EMG signals (muscular activation), ultrasound images (muscular shape change/stiffness), and respiratory data (energetic cost). We will apply this collected information to optimize controllers to drive specific changes to the musculoskeletal system as well as assist/enhance locomotion in specific populations (elderly, stroke patients, military, etc.). We are currently looking for students from any engineering discipline who are interested in and can assist us with data collection and analysis, controller implementation with a user interface (MatLab), controller optimization and machine learning, and musculoskeletal modelling. Contact: Ben Shafer (Ben.Shafer@gatech.edu).
PEDIATRIC KNEE EXOSKELETON
Description: The team has built a knee exoskeleton specifically for children with walking disability mainly caused by the knee joint such as cerebral palsy. The idea of the project is that the device partially assists the user's knee joint effort during walking rehabilitation. The team will focus on running experiments with children and experimental data analysis. The experiment with children as well as with able-bodied adults would provide a valuable experience in human-subject testing. The data that will be collected from the experiment include 3-D motion capture, energetics, biomechanics, muscle activity, device performance, and user’s feedback. Using the data, biomechanical analyses will be performed to evaluate the effectiveness of the powered assistance from the device. Students from BME, ME, and ECE backgrounds are all suitable for the project. Additionally, students with skills/knowledge in machining, LabVIEW, Matlab, circuit design, or biomechanics will be helpful. Contact: Dawit Lee (email@example.com)
Description: This team is dedicated to developing and validating high-level controllers on a knee and ankle prosthetic device. The main goal involves creating a user independent system to allow persons with transfemoral amputation to improve their ambulation in common community tasks. We will be collecting data from different sensors which included embedded mechanical sensors (IMU’s and encoders), EMG signals, motion capture, and metabolics on a variety of terrain. Skills that are desired but not limited to include programming skills (C, C++ & Python), mechatronics (PCB design & embedded systems design), and strong mechanical design background (SolidWorks). We are looking for motivated students from any discipline. Contact: Krishan Bhakta (firstname.lastname@example.org)
ROBOTIC HIP EXOSKELETON
Description: The hip exoskeleton team will primarily focus on able bodied subject testing in the terrain park. We will test in multiple inclination levels and provide power assistance to the user to observe the exoskeleton’s effect on human biomechanics. Additionally, we will work closely with patients with gait deficiencies using our hip exoskeletons. We will collect multiple data sets from different sensors/robots such as muscle signals, motion capture, human energetics, and the device data (encoder, IMU etc). Students participating in the project will help the team in their own specialties, such as biomechanical analysis using Vicon motion capture and OpenSim modeling/simulation software, mechatronic design/fabrication and control of the new actuator, and integrating machine learning techniques to sensor signals for controlling the exoskeleton. As the project itself is heavily interdisciplinary, student will be exposed to diverse mechanical, electrical, and biomedical skillsets from controlling robots to working with clinical populations. Contact: Jeff Hsu (email@example.com) and Inseung Kang
Methods and Technologies
Academic Majors of Interest
Preferred Interests and Preparation
As the research is an interdisciplinary research, wide range of skills and experiences will be highly applicable to the study. Some of the core skills will include mechanical design, machining, mechatronics, embedded and MATLAB programming, and signal processing. Below are examples of specific skills that would be related to each major.
ME - Background/interest in mechanical design and machining for fast prototyping. Training in Montgomery machining mall for CNC, mill, lathe, 3D printer is recommended.
EE/CompE - Background/interest in mechatronics and electrical circuitry design. Knowledge in circuit design and signal processing is recommended
CS - Background/interest in software development for embedded programming. Knowledge/interest in language (C, Labview, MATLAB, python) is recommended.
BME - Background/interest in clinical testing for device evaluation. Knowledge/interest in human biomechanics and design of biomedical robotic devices/controllers as human and biological data processing is recommended.
Meeting Schedule & Location
- Mechanical Engineering