PURPOSE

Development of a device to offer augmented strength and stability for individuals with weakness and the occurrence of tremors in their hands, as experienced by sufferers of Parkinson's Disease 

 

RESULTS

Fully functional robotic exoskeleton prototype, comprised of an aluminum linkage structure powered by pressure sensing, servo actuated, closed loop motor control system. 

Parkinson's Disease afflicts tens of millions of people around the world, leaving many with, among other symptoms, limited use of their hands due to tremors and weakness/atrophy. Beyond this, there are countless other individuals affected by various forms of functional impairment in their hands. Taking this into account, and driven by the desire to have some positive social effect, my five member Mechatronics class project team decided to see if we could design a device that used our mechanical engineering skills to address these issues.

After extensive initial ideation, our group decided to design and build a robotic hand exoskeleton that would use some form of closed-loop signal processing to control hand motion, assisting when it is desired and restricting when it is not.  The primary reasons for pursuing this type of device were as follows:

  • Exoskeleton structure could simultaneously provide assisting force to hand as well as restrain undesired motion
  • Proof of concept for mechanical design could be applied to any device where augmented strength was desired, largely independent of specific cause or parameters
  • Proof of concept for control system could be applied to multiple devices where signal manipulation/filtration was required, largely independent of hardware

The design process was divided into two sub-processes, mechanical design and control design. As mechanical design lead, I was able to create and oversee the following:

  • Initial hand sketching and mechanical ideation
  • CAD (Solidworks, PTC Creo) based 3D modeling
  • Motion and force analysis
  • Systems layout and integration 

In addition, I was able to assist with developing the control architecture, comprised of:

  • Piezoelectric pressure sensors
  • Arduino micro-controller
  • High torque, metal gear servo motors

The design consisted of pressure sensors located inside cuffs attached to servo driven linkages, with users placing their fingers inside the cuffs. When users try to open or close their fingers, the sensors pick up changes in pressure, actuating the servos to offer assisting force. A program was written for the controller that filtered out pressure sensor signals occurring above a certain frequency, as would occur during tremors. The rigid structure would also work to contain tremor movement, leaving room for only smooth motion along a desired path to occur.

design glove.png
build glove.jpg

In addition to user functionality, an integral concern was designing in a manner that would ensure speed, flexibility, precision, and accuracy during the manufacturing process. To this end, I geared all CAD modeling and manufacturing planning around advanced, highly automated fabrication techniques, including:

  • water jetting 
  • laser cutting
  • 3D printing

This approach allowed us to maintain a high level of confidence with respect to our CAD motion and force analyses. In addition, post-automated manufacturing processes were limited to basic assembly, removing time consuming and risky hand fabrication techniques.

glove team.jpg

Once all components were assembled, the control system was tested. After initial calibration, the device successfully offered high torque assistance for smooth opening and closing of fingers while neglecting any high frequency pressure signals and not allowing any shaky, tremor like motions to occur.

There were several lessons learned during the process. While functional, the piezoelectric sensors exhibited signal noise that had to be addressed in the programming. A different type of sensor, possibly capacitive, may have been more effective. In addition, current limits in resolution with certain manufacturing techniques became apparent, such as the inability to water jet functional gears below a certain dimensional threshold.

With future development and greater time and manufacturing resources, it seems the mechanical architecture could be downsized and made lighter for those using it in a medical capacity. Inversely, a larger, more robust structure could be developed for other potential uses, such as industrial applications where greater-than-human hand strength might be desirable.