Robotic Arm Rehabilitation Device

Assisting Tecnalia's venture in designing a rehabilitation robot to aid stroke victims in their recovery. Our aim is to further improve Team Arm Assist's two previous add-on modules by implementing advancements to create a more fluid, functional product.

Development and Project Goals
TECNALIA's Arm Assist Robot is an assistive robot for the rehabilitation of patients with upper limb neuromuscular impairment, specifically for patients who have suffered from a stroke. This assistive robot is accompanied by a tele-rehabilitation platform, thus allowing rehabilitation at home under supervision from the therapist by its quantitative evaluation.

ARMASSIST represents a major technological breakthrough which allows TECNALIA to address the growing market demand for personalized care. This initiative was born to meet the needs of the 15 million people who suffer brain-vascular damage every year; of that figure 5 million suffer permanent disability.[]

The Robotic Arm Assist team, provided by the University of Idaho, will work through the capstone project to further develop the rehabilitation robot. The team will add to the overall functionality of the robot by providing TECNALIA with modules that will increase the ability to rotate the wrist and ability to open and close the patient's hand.

Problem Definition
Our goal is to evaluate the 2013 Team Arm Assist's modules and make further improvements. We are primarily focused on the advancing the functionality of the wrist and grip module to create a universal product that can be powered comfortably by any user. Our improved modules should be durable, increase adjustability, and ensure accurate results to guide the patient to recovery.

Background


Our team, Team Rehab, is continuing the Robotic Arm Assist project from the 2013 team.[] Team Arm Assist designed three add on modules that consisted of a grip, wrist rotation, and arm elevation module. At Tecnalia's request, they want us to further implement advancements to the grip and wrist rotation module while the company will be finding a more efficient design for the arm elevation module.

Last year's prototype lacked in a few areas that need further development. Aside from the lift module, the main areas that can be increased within the grip module include adjustability to accustom all hand sizes, implement a sensor to measure force and degrees in grasp, and create an adjustable system to change the resistance. The previous year's sensor gathered poor and inconsistent data. Therefore, it is our goal to design an efficient method to test the force and degrees of grasp so that the data will be consistent and concise with every user. To the right, is the final design of the grip module that we will improve.



In the wrist module, the primary characteristics to be improved include incorporating a locking mechanism to prevent pronosupination, add a gyroscope to calculate degrees of rotation, and design soft stop ends at the end of rotation. The picture on the right includes last year's design. The blue acrylic piece in the center is where the grip module mounts and is then allowed to freely rotate via two aluminum rails. The stops on either end are hard plastic that abruptly stops pronosupination of the wrist which is sought out to become absorbed from a design by this year's team.

Deliverables
When determining the necessity of each improvement that was specified by Tecnalia, we assembled a metric matrix that listed all the needs and the level of priority. Below is the following list that decided where we need to focus and put most of our effort into to fully satisfy our client. The scale in which the priority was rated by Tecnalia was on a scale from 1 to 5. 5 - Must have. 3 - Desirable. 1 - Luxury.

Specifications
 Grip Module 
 * Measure force of grasping and achieve accurate/consistent feedback
 * Make universal by providing adjustability to accustom all users
 * Provide comfort and simulate natural human biomechanics of the hand

 Wrist Module 
 * Allow full range of motion in wrist (±90°ideally)
 * Incorporate smoother stops at end of pronosupination
 * Create a locking mechanism to prevent movement in wrist
 * Integrate calibration system to record the degrees of rotation.

Project Learning
The process of implementing new improvements at the request of Tecnalia has entailed a lot of research from the previous year's results. There was a learning period where we had to quickly become up to date where Team Arm Assist (2013)[] had finished. This included heavily examining their project portfolio to see where they succeeded and where there was potential improvement. Since their final add on modules were in Spain, we had to fully recreate the wrist and grip module from scratch. This entailed understanding Team Arm Assist's assembly plan, function of each mechanical part, and where there were was error in final calculations.

Since we had to recreate their previous modules, we were able to become proficient with the machine shop tools and the 3D printer MakerBot. As we have constructed and machined parts from their final design, there has already been areas where we have found that can be simplified and further improved.

Assembly Process

Machine Shop


 * To speed up machining time:
 * Reduce tool changes
 * Avoid re-zeroing- there are also faster (less accurate) techniques to zero such as running the cutting bit on to just barely cut the surface for non-critical dimensions
 * The attachment technique of the part is critical for the machining process- e.g. for press fit parts, the tolerances are very tight
 * 3-d printed parts have bad tolerances (done here) so you may have to adjust dimensions to fit the printed part

Grip Module


 * Thumb piece is difficult to attach and becomes loose after a few minutes of use.
 * The pin used to adjust the length is too small and breaks easily.
 * Springs on the thumb piece rub against the sides of the part.
 * It has to be screwed into the pronosupination module and cannot be used separately.
 * The grip has a rigid plastic on the back of the hand which limits adjustability and comfort for patients.

Wrist Module


 * Grip module rubs against the box used to mount the pronosupination module
 * The rings takes up a lot of space and could be reduced in size
 * Certain components of the base could be combined to reduce the assembly process time and increase simplicity.
 * The stops at the end of the rails need improvement to make an easier and smoother stopping motion at the end of pronosupination.
 * Needs to improve design and create more space for sensors within the base.

Grip Module

During the creation of last year’s grip module, it became apparent the thumb piece will need to be redesigned. The piece was uncomfortable and the springs attached to it rubbed against its plastic walls when in use. Because of this the springs do not move in a linear motion. The grip module also needs to be redesigned to make it more adjustable for different hand sizes. The current version can be adjusted length wise but the pieces used to do so are weak and can break. We are trying to make a new module that is easily adjustable for both hand length and width. Currently we are looking at several different designs taking inspiration from adjustable exercise equipment and gloves designed for people with poor hand movement. In the future we may make the grip module out of carbon fiber since it is a stronger more reliable material than plastic.

Wrist Module

Overall, the wrist module does not have any substantial or demanding areas of improvement. The swivel motion is natural when pronosupinating the wrist because the rails and bearings enable smooth motion.

Sensors

To provide real time data on motion and positioning, several sensors along with a dedicated microcontroller have been integrated into the modules and ArmAssist system. A quick search reveals a wide variety of sensors that monitor different types of motion. Implementing the correct sensor required extensive research into each sensors capabilities. The result is two rotary encoder sensors in the grip module to measure angular position and speed, along with a gyroscope in the wrist module capable of reporting orientation and measuring rotation of the module. These sensors are monitored and read by a Beaglebone microcontroller.


 * Learn more: Rotary Encoder []


 * Learn more: Gyroscope []


 * Learn more: Beaglebone []

Consultants
  Dr. Joel Perry 

Project Manager - TECNALIA, Rehabilitaiton Technologies Dept., Health Division

Adjunct Faculty - University of Idaho, Mechanical Engineering

  Samuel Qualls 

Fall Semester Graduate Student Mentor

Team Documents
Drop Box Link: []