Human Safety: Driven by Artificial Intelligence

August 1, 2019

By: Riley Noland 

 

TuMeke- https://www.tumeke.io/

The Safety Industry:

Safety is the most important part of a company’s culture. The traditional approach to ergonomic safety is hard to scale- it requires a team of specialists to implement correctly. While wearable technologies offer some solutions to this issue, they come packaged with their own set of headaches: excessive costs, intrusive design structures, etc. So does there exist a low-cost tool combining both the big data benefits of wearable technologies with the unintrusive benefits of a human observer?

The Solution:

Yup! Recent advances in Artificial Intelligence and Computer Vision have given us the ability to quickly analyze human motions and automatically compute the risk of a recorded job. That is why we built TuMeke: a non-intrusive mobile application to help safety professionals measure and track the health of employees.

Simply take a video of an employee doing a task. TuMeke will store your video and produce a 3D model of that person doing the task to show which parts of the recorded job are at risk of causing injury. From there, you can easily navigate and compare your videos to help train and improve your workplace’s ergonomic performance.

Building this application has been a challenge in of itself. Creating a neural network that is dynamic enough to analyze human motions through only computer vision and using that to compute the risk of a job in under 4 minutes is no simple task. But we know technology has advanced such that it is possible, the question was simply who was willing to spend the time to devote it to ergonomic safety.


   image 1image 2  image

 

 

 

 

 

 

 

 

 

 


The Burden:

Every year, billions of dollars are spent helping those who were injured while on the job. In lifting alone, OSHA reports that the direct cost of lifting related injuries in the US are around $15 billion. However, that number spikes to $56 billion when we take both indirect and direct costs of lifting related injuries into account . By using TuMeke, anyone at a firm will be able to proactively isolate and stop damaging motions that could result in costly recoveries by simply taking a video.

The Team:

Our team is small, young, and really excited to work with you! Two of the three founders are undergraduate students at Stanford University studying computer science and mathematical logic. Our investor is one of Silicon Valley’s premier Early Stage venture capital firms, OVO Fund. Reach us at https://www.tumeke.io/ if you would like to demo TuMeke or learn more about what we’re doing!

The Journey:

TuMeke began as one of many entries in a Stanford University business pitch challenge. The initial idea involved using our joint tracking software to help gym goers recognize good and bad weight lifting form. As we further developed TuMeke though, we realized that a much more important issue existed within industrial settings. More specifically, we realized that TuMeke could significantly improve individuals’ lives by helping prevent severe musculoskeletal injuries.

TuMeke is still in its infancy, but with your help we know we can tackle injuries head-on. We are looking to have a beta version of our software out within the coming few weeks. To request a demo visit us at https://www.tumeke.io/!

We look forward to working with you.

Author Bio:

Riley is a senior at Stanford University studying Symbolic Systems with a concentration in mathematical logic. In his free time, Riley enjoys hiking around the San Francisco Bay Area or going on cross-country drives with his friends.