Safety and Artificial Intelligence

There’s a lot of chatter here on LinkedIn and other business social media channels about artificial intelligence, how much those with AI skills can earn, and robotics. All of this information can be overwhelming and it can cause a person to push back saying it’s too futuristic.

The future is here, now, and you’d better get educated! For my fellow safety pros, there’s so much opportunity to get in at the ground floor in our organizations with using AI and robotics. This post is for safety pros to begin their study of AI and robotics so you can talk intelligently about the topic and be a leader at your organization.

The lowest hanging fruit is fear. No matter the organization or how techy the individuals, there will be fear that robots will kill us. It’s something my favorite podcaster Gary Vaynerchuk casually mentions all the time.

The International Organization for Standardization (ISO for short), published a Technical Specification (TS) in early 2016 about working with “collaborative robots.” It’s ISO/TS 15066:2016 and it’ll cost you about $140 to download or receive a paper copy. The TS and related article can help you begin the conversation with personnel at your company about fear of robots.

This article is a great start to begin answering the question “how can we work safely alongside robots?” The challenge is programming AI to understand safety. Think about how hard it is to “program” safety into a human worker and you can see the challenge ahead for AI developers. The Safety Principle is part of 23 principles for AI developed at the 2017 Asilomar AI Conference. The 23 principles are supported by Steven Hawking and Elon Musk as part of 1200 AI/Robotics researchers and 2342 others. You can add your name to the support list by going to the previous link.

The Safety Principle is the first listed under the heading of Ethics and Values and reads as follows –

Safety: AI systems should be safe and secure throughout their operational lifetime, and verifiably so where applicable and feasible.

What about a worker that’s part robot and part human? That’s HAL (Hybrid Assistive Limb) in the photo at the left, from Japan – and it’s meant to assist human physical functions for workers, elderly, and those with nerve and muscular disorders.

The next logical way to approach the AI and robotics topic at work is injury reduction. No matter the industry, soft tissue injuries like sprains and strains are a common source of injury, often resulting in costly and lengthy workers compensation claims. It makes sense to look to technology to reduce this type of claim. Contrary to what some fearful people at your organization will say, robots will not completely replace humans! I believe that there will always be work that will involve a human touch, and that work can, and should, be made more efficient with technology.

Wearable exoskeletons are currently in use in several industrial settings and construction sites. NIOSH published a blog in early 2016 that has many links for further study. Here’s a post from NIOSH from this summer about exoskeleton use in construction.

The last approach I’ll discuss here is motor vehicle safety. Telematics and other monitoring applications are already in use with most large fleets. These systems enable managers to know where vehicles are, how fast they are going, and receive exception reports when braking is too sudden, cornering too fast, or a vehicle is outside of its intended geofenced range. You’d have to be living under a large rock to not know about the push for self-driving vehicles. They’re already being used by consumers – take a look at Tesla’s AutoPilot system for their Model S and X vehicles.

I’ve personally driven a Tesla Model X with autopilot and this article from Business Insider echoes my feelings on it being both awesome and terrifying!

The foundation of self-driving vehicles is lidar – short for “light detection and ranging.” You can learn more about what self-driving cars “see” by reading this post from the New York Times. If your company is looking into expanding their fleet with vehicles that have self-driving capability, it’s important to evaluate the lidar system the vehicles have – not all systems are created equal.

What do you think? Have you already started your AI education? Let's talk.

This post originally appeared at www.theferrigroup.co - where you'll find more information on safety and technology presented in a quick and relevant manner via blogs, podcasts, and infographics.