The Armature Project

Votes: 12
Views: 3203

Problem:
In 2017, the Global Burden of Disease (GBD) declared that among the 57.7 million amputees in the world, 19.6% of the global amputee population is suffering from unilateral upper limb amputation while 19.1% were bilateral amputees. Subsequently, the market for active upper limb prosthetics is divided into two types, namely prosthetics using pure EMG and prosthetics based on computer vision in combination with EMG. All of which rely on the significant change in EMG readings to control their gesture recognition. According to a study done by Cheesborough et. al not all upper limb amputees can produce significant muscle activity from their residual limb. While, Kukien et.al have investigated the possibility of remapping the original EMG location from the residual limb to different parts of the body such as pectoralis major and the pectoralis minor. However, their remapping method creates an unintuitive control system.

Solution :
Instead of remapping the EMG location for upper limb active prosthetics, we propose a prosthetic arm called AI Armature that utilizes artificial intelligence and computer vision via smart glasses. It makes use of a gesture control system to circumvent the need of EMG sensors to control the overall system. The proposed system reduces the effort needed by upper limb amputees to actuate their prosthetic hand. The computer vision system will rely on eye tracking and object detection to assist the gesture predicted by the Yolo v4 model. To reduce erroneous object predictions, we propose the use of a surjective object-gesture mapping approach to increase prediction accuracy.

Market Potential:
With the global market amounting to 38.7% of the total amputee population, which roughly amounts to 22.3 million amputees, we believe that there is a market for our AI prosthetic arm. The upper limb prosthetic market is evaluated at 1.08 billion USD back in 2018 and is forecasted to grow up to 11% in sales from 2019 to 2025 according to Ankita Bhutani et. al. By 2025, the study projects that the market value will be evaluated to be at least 2.3 billion USD. In this period of growth, the AI Armature team plans to be one of the leading upper limb prosthetic manufacturers in the world.

Partners:
The AI Armature project was able to secure NT$ 2 million, in this year’s recently concluded FITI (From IP to IPO) competition held in Taiwan last June 18, 2021. Furthermore, we are partners with Mingder Medical Orthotics for the OEM prosthetic arm manufacturing, Jorjin for the smart glasses and AverMedia for AI device. With support from the local doctors, amputees and businessmen, we are set to improve the lives of amputees.

Design:
We reduce the production cost of the arm by making use of concurrent 3D printing methods. On top of optimized 3D printing techniques, we plan to package the overall system with already available smart glasses and AI devices. Furthermore, we have minimal motors to reduce power consumption and reduce weight. Lastly, we are continuously innovating to reduce weight and increase grip strength.

Video

  • Awards

  • 2021 Manufacturing/Robotics/Automation Honorable Mention
  • 2021 Top 100 Entries

Voting

Voting is closed!

  • ABOUT THE ENTRANT

  • Name:
    Jolo Tolentino
  • Type of entry:
    team
    Team members:
    Cheng, Jia-Xiang
    Hu, Jin Teng
    Mo, Yu-Kuan
    Policar, Roberto Jacob
    Tolentino, Jolo Gerard Miel
    Prof. Chang, Wan Jung
    Prof. Chen, Ming-Che
    Prof. Ou, Yang-Kun
    Prof. See , Aaron Raymond
  • Profession:
    Student
  • Patent status:
    pending