
Using AI for Safety
Is anyone using AI at their facilities to help comply identify near misses? I attended a webinar yesterday for Protex and their software integrates with your existing security system to auto identify things like speeding forklifts, PPE not worn, PIT/pedestrian interactions, etc. It looked really cool in concept but I haven't heard of anyone using it yet.
I pressed them for pricing and for 1 location with 10-20 cameras the cost was $40-45k/year. Seems expensive but if it helps be the eyes when no one is watching, maybe it would be helpful?
Here's a 2 minute demo if anyone else is interested:
https://www.protex.ai/watch-demo
Comments (4)

I imagine the only ones using something like this are probably fortune 500 companies....assuming more adoption as pricing comes down over time
Really great product, and a lot of companies starting to incorporate AI Vision systems
I don't support the use of AI as the next phase of safety compliance. I'll be honest and i'm not directing this at you, just sharing my thoughts. It's Orwellian. Safety is not watching everyone 24/7, sending the infraction notices in real time. It's totalitarian and contrary to what real safety is. It's people. If the company really cares, it invests in it's management, its people. That's what culture is. AI based solutions for continuous monitoring will only put a wider gulf between the worker and management with us (safety people) smoothing it over. At some point, after training and resources are in place, you have to trust people. It may have it's place somewhere for investigations, but i think IF it turns into something affordable and isn't a fad (right now it's just a fad) it could be useful. But, we live in an age where people want total control for the sake of profit...not trying to be political but it's the truth. Just because we can, doesn't mean we should. Should we put breath test units on every car you have to blow to drive to prevent drunk driving? Should we put "return to the dealership" tech on cars if a working family hits a road bump and misses a payment? We need to stop considering the cost of AI and start with considering it's human cost and whether it's right at all.