AI-Driven prosthesis for the planet | Enzo Romero | TEDxCalle Aldama

Quick Overview

Enzo Romero, founder of Latbionics, advocates for making AI-driven prosthetic technology accessible to low-income individuals, highlighting that while their current prosthetic success rate is 50%, the ultimate goal is to restore dignity by ensuring access is not determined by income or geography.

Key Points: Enzo Romero founded Latbionics to develop affordable, AI-driven prosthetic devices for low-income individuals in Peru and across Latin America. The current success rate for users integrating the body-powered prosthetics is 50%, which Romero finds alarmingly low. The AI-driven electronic prostheses rely on myoelectric signals detected by sensors placed on the user's residual limb muscles to infer intention for movements like wrist flexion/extension. The entire process, from 3D scanning the residual area to designing the prosthesis on a computer, is done locally. Romero's goal is that dignity, not income or geography, determines access to assistive technology. The organization aims to increase the success rate of the AI algorithms from 50% to 97% accuracy.

Context: This TEDx Calle Aldama talk features Enzo Romero, a mechatronics engineer and university professor, who discusses his work with Latbionics, an organization focused on creating accessible, affordable prosthetic limbs using digital manufacturing technologies like 3D scanning and printing. Romero shares his personal motivation, stemming from his own experience using assistive technology, and outlines the challenges and progress in deploying their AI-driven solutions in Latin America.

Detailed Analysis

Enzo Romero, founder of Latbionics, presents his organization's mission to provide AI-driven prostheses to low-income populations in Peru and Latin America, driven by his personal belief that dignity should not be limited by economic status or location. He begins by questioning the audience about their experience with assistive technology, noting that 1 in 7 people globally have some form of disability requiring such devices. Romero reveals that his childhood dream was to have a prosthesis like Luke Skywalker's. Latbionics develops two types of prostheses: body-powered ones (which operate via residual limb movement) and electronic ones (which use sensors placed on residual muscles to read electrical signals indicating intended movement, like wrist flexion or extension). He shows an image of the 3D scanning process and the CAD design phase using software like Autodesk Inventor. The current success rate for users successfully operating the body-powered prostheses is only 50%, which he considers too low, noting that users often stop using them due to fatigue. The goal of their AI algorithms is to learn these muscle signals to infer user intent, aiming to raise the accuracy rate to 97% for the electronic prostheses. Romero emphasizes that innovation must be driven locally and committed to dignity, rather than being constrained by user income or geography.