Q&A: Self-service AI tech for medical device repair
Med-Tech Innovation News spoke to Eitan Cohen, CEO of TechSee, a company which has been offering medical device repairs virtually without the need of an in-person technician.
First of all, tell us about TechSee and its role within the medical device sector?
TechSee delivers visual assistance powered by video, AR and computer vision. This technology can be used across several use cases in the medical sector:
- Medical devices (consumer) – remote guidance for patients regarding how to set up, operate and troubleshoot their medical devices. For example, we work with Coachlear to guide their users on how to set up and use their hearing aids (who cannot get a traditional phone assistance). This remote service reduces costly device returns, enhances customer experience and also supports upsells.
- Medical devices (B2B) – enabling remote maintenance and repairs of medical devices such as ventilators, MRI, IPL etc. Hospital staff can simply point their smartphone or tablet camera at a medical device and our Computer Vision AI can instantly recognise its make, model, components, wiring, LED lights, cables, ports, and circuitry. This allows them to repair delicate equipment either with the help of a remote professional or entirely on their own, with AI-driven self-service. Remote service is crucial to avoid risk of exposure (e.g. supporting equipment in COVID-19 hospital departments) as well as saving costly technician visits and machine downtime.
Our technology is used both by consumers and medical technicians, in ‘assisted mode’ (where a human expert guides remotely) or in automated self-service where the user can simply point his smartphone camera and interact with AR Assistant that can see. This is based on our computer vision technology (learning to recognize specific devices, their models, parts and identifying their issues).
Is it a simple case of remote repairs?
We support both simple and complex cases. Examples:
Simple cases: Repetitive customer calls around glucose monitors, hearing aids, and other at-home medical devices. The customer can simply use their smartphone or tablet camera and point it at the device and be guided to resolution, either in fully automated self-service mode or with the help of a remote expert.
Complex cases: A medical technician has wired an X-Ray machine and wants to confirm that the job was properly done. Instead of calling a remote supervisor, he can simply show the job to our Computer Vision AI, which can provide verification that the job was done correctly.
How complex a repair can you conduct for a medical device?
We believe that computer vision AI can handle 80% of repetitive cases, where the 20% of the most complex cases should still be handled by a human expert. Our technology has reached 98% accuracy of devices and issue recognition in complex equipment.
How closely do you work with manufacturers to build your knowledge base?
Each manufacturer can leverage their visual data to create computer vision models. This visual data can be collected during video interactions with customers, images taken by medical technicians, or created in the lab (synthetic data), in a way that enables that organisation to scale and train new models in a matter of days.