Seating Coach

My goal is to help classify the propulsion patterns of wheelchair users into four categories: semicircular, single looping, double looping, and arcing. I am currently utilizing the accelerometer on the eWatch in conjunction with an accelerometer on the wheelchair frame to collect data on propulsion patterns. The propulsion patterns will be determined by extracting features such as mean and variance from the acceleration data. We also plan to develop a virtual coach system. The virtual coach will prompt the wheelchair user to change propulsion patterns depending on the environment in order to reduce the possibility of injury.

Auto Dock

I'm working on the Autodock project. Autodock is a system we hope to create that allows wheelchair users to be safely fastened in a public bus in one-third of the time it currently takes to fasten someone in a wheelchair

Robotics Therapy

I am analyzing hand rehabilitation and robotics; designing hand rehabilitation exercises to be implemented on robots; using OpenGL to create and implement graphics for target exercise, and incorporating robots into target exercise.

Integration of an Electrical Powered Wheelchair and a Robotic Arm for Coordinated Manipulation

This summer I am assisting in the design, modeling, and fabrication of the necessary components to interface two commercially available robotic arms with a Permobil C500 electric powered wheelchair. The long term goal of this project is a virtually autonomous wheelchair that uses a variety of sensors to allow a person to safely and successfully navigate any environment or situation. For example, if the operator needs to open a door, he or she would indicate this by perhaps aiming a laser or camera at the door knob which would then signal the wheelchair to open the door through coordinated movement of the wheelchair and arms. A majority of my time here this summer will be focused on designing and creating of a track on three sides of the wheelchair seat that will add an additional degree of freedom to the robotic arm. This track will allow each arm to travel to both sides of the wheelchair and also permit the arms to be tucked behind the wheelchair for assess through tight opening. This project includes extensive mechanical design and machining and some integration of electrical components.

Scientific Overview of Knee Osteoarthritis

My research investigates common rehabilitation programs for knee osteoarthritis patients and their common criticisms.  Current technology addresses some of the issues preventing patients from reaching maximum functional capacity. Future technology is being developed to increase exercise adherence in individuals with knee osteoarthritis.  The development of a robotic coach that is aware of exercise performance and encourages exercise adherence may overcome possible barriers.  This social robot addresses cost efficiency, self-efficacy, motivation, and self-monitoring of patient rehabilitation.

Safe Driving

Driver safety has become an important issue as national businesses expand and increasingly rely on fast shipping to transport products cross-country. Several methods of monitoring driving habits such as driver-vehicle interaction have been developed using SICK laser scanners. However the practicality of using such devices for mass consumer distribution is not feasible because of the expense. An alternate option for censoring devices is a technology that has been around for quite a while, Sonar. Although Sonar is far less accurate than laser scanning, it cost only a fraction of what a laser scanner would cost. In applications where a vehicle is in close proximity to another object, such as parking and low speed monitoring, a Sonar sensor would be sufficient in place of a laser scanner. My project is to use data collected from the SICK lasers and use it to determine how much of the original data would be detected from the Sonar using sound wave propagation properties. The simulation is compared then to a physical Sonar, to test the accuracy of the computer method.

Memory Coach

I'm working with Dr. Anind Dey and the graduate student Matthew Lee at the QoLT, in the Human Computer Interaction Institute, Carnegie Mellon University. The project that I'm involved in is research for giving memory support to Alzheimer's patients and their caregivers. Our goal is to develop a software/hardware system that will support the needs of both sufferers and caregivers, in helping them to recall recent experiences. The system will capture the past experiences (pictures, audio or other multimedia) along with appropriate tags/cues so that later on the patient and the caregiver can review these events for the purpose of   improving the patient memory abilities. Because a large number of photos and multimedia can be taken in a day, we need to develop software that can pick up just the interesting things.

Inside out Vision

I am building an independent sensor system that can record many instruments placed on various locations of the human body. These instruments include web cam, microphone, accelerometers, gyroscopes, blood pressure  sensors. The purpose of this is to create a database of all possible information that can be captured from a person performing any task in a first person view. In addition to this an environment will be created in order to record not only first person but also third person view.

Mobility Aides for Persons with Multiple Sclerosis

Ana’s study: We are collecting data to determine the most appropriate assistive devices for people affected by Multiple Sclerosis. Eligible study participants are given questionnaires that evaluate both the participant’s current condition in regards to the disease and the participant’s current satisfaction with their assistive device. Once enough data is collected and entered in Microsoft Access, we will compile this information and draw parallels b/t the way MS affects a person and the type of assistive device that would give them the most satisfaction with their quality of life.

Hannah’s study: Disparity in Wheelchair Quality of Veterans with Spinal Cord Injury
We are collecting data in order to explore the relationship b/t quality of wheelchair and several demographics. There are multiple hypotheses we will test using this data. For example, are the wheelchairs supplied/ funded by private insurance of a better quality than the wheelchairs supplied by the VA system? Eligible study participants will be mailed questionnaires and the data will be entered in Microsoft Access.

Remote Wheelchair Logging-Analysis-Safety System

We are building a wheelchair Logging-Analysis-Safety System capable of recording wheelchair usage, analyzing contexts, and communicating with the host station via cellular communication. This project will entail the design and development of hardware and software that will serve as a platform for further improvement and investigation of wheelchair usage and activity, and community participation of wheelchair users. I am working toward three project objectives:
ndash; To build a prototype of the remote wheelchair logging-analysis-safety system that integrates two sub-devices, i.e. a device worn by the user, and a device onboard the wheelchair
– to implement cellular communication between the system and the host computer for data transmission and safety service in atypical conditions.
– to implement a real-time, wheelchair-based context-aware system capable of classifying sensor data to detect locations and wheelchair failures.

Exercise Coach

I am working on two projects, one on motivating physical activity and the other on Instant Messenger power usage. The physical activity project uses a pedometer, cell phone technology with an accelerometer, and a GPS unit to analyze people’s levels of physical activity. This data is then uploaded to a web page where people can view how physically active they are being, as well as where they are and who they are with when they are being the most physically active. The power usage project is one that looks at AIM users on a laptop, with an IMBot that can record when users are idle and when they are active. Based on this information, along with information supplied by the user about laptop power settings, one can tell how much power is being wasted when AIM users are idle but not closing their laptop or putting it to sleep. In both of these projects, I am looking at different types of feedback mechanisms and figuring out which types work best to motivate behavior changes.

Carnegie Mellon University
Freshman
Major: Mechanical Engineering
Faculty Advisor: Anind Dey
Student Mentor: Ian Li