Interview with the Wire Journal International: How we use AI for safety
Before we ask about what you do with AI, what sets you apart from other companies with similar technology?
What differentiates us from the other companies out there is that we are solely focused on EHS. Our AI platform processes 300,000 images daily and to date has processed more than 16.5 million hours of visual data in total and still counting. The platform provides you with 40+ pre-trained use-cases developed through the interviews made with requests and feedback gathered from clients in various industries. Yet, the platform is flexible enough to anticipate the industry-specific requirements. The Intenseye team analyzes and alters the use-cases according to the customer’s needs as much as possible. Our customers pay no extra fees for development efforts and lose no time with long analysis sessions. It’s just implementing, setting up and starting EHS compliance and analysis in hours.
How does your AI technology track multiple employees, and what exactly is it seeking and how does it know what to look for?
We get connected to the existing CCTV cameras in our clients’ facilities very quickly. We have 40+ enterprise-ready, proven to plug and play use cases with high accuracy levels. Our use-cases are all employee health and safety-related focusing on high-risk activities that are leading to life-altering events, including fatalities. Our customer success team guides our clients as part of our onboarding pack and helps to set the rules together with our customers. Setting up rules is very simple in the form of a drag and drop. After a few hours of induction, our clients are set to go!
Do you have a basic program that works for any manufacturing operation that you adjust for different sectors? How difficult is it to do that for wire and cable manufacturing? Are you able to offer this now?
One great news here is that we are industry and sector agnostic. We operate in multiple sectors, ranging from manufacturing to logistics, from oil and gas to retail. So, when I look at the wire and cable manufacturing, the vast majority of the EHS related risk factors are around getting stuck in between moving objects, being hit by moving objects, working at heights, restricted work locations, walking under suspended loads and ergonomics. All these use cases are available in our core AI pack.
What happens when the AI sees a potentially dangerous action? Who gets notified, can such notifications ever be used against the company if there's an accident?
It is basically up to the users of our platform, however they wish the follow-up to be. You can get the notification on your mobile phone as an SMS, or email or just get it on the platform as a data point that will be reported to you on a regular basis. Additionally, since the AI is working 24/7 in real time, you can integrate alerts with the smart devices in the field as well. For instance, you can hook it up to an IP loudspeaker, or an IP visual alarm so that you can get the workers notified in real time as well. Those notifications, in our experience so far, are not being used against a company. In fact, we as a company seek to empower the workforce to help them save their own lives.
Can you describe the actual results from the use of your AI? Did getting notifications of unsafe acts & conditions result in changes in employee behavior? Is there potential for reduced workers comp insurance rates?
Intenseye sees the “unseen” 24/7. When Intenseye captures unsafe conditions, we encourage our clients to talk to the workers and teams at the frontline, openly. It is so crucial to invite the frontline workforce to talk about the reasons why the unsafe act or situation occurs. As a company, we strongly believe that “context drives the behavior.” Thus, we steer the leadership of our clients to address the systemic level issues, to be able to provide the right context where people can work safely. And indeed the incident rates are going down and the compliance rates increase over a short time period. One of our successful customers picked the right priorities thanks to the insights from Intenseye and identified the systemic deficiencies that cause injuries related to manual handling. They worked collectively with their frontline workers to fix these issues and made compliance easier for them. We observed an improvement in the count of unsafe alerts by 65%.
Has your use of AI evolved since it was first deployed? Does it evolve as there is more experience in a given field? Can a single motion be interpreted by AI as a danger when it may not be, and could there be hundreds of such violations that could overwhelm a company?
AI has evolved hugely since first deployed, actually we have been growing based on the requests from our customers. We turn the newly captured data into training data so that the AI is being fed with more data continuously. I’d like to emphasize the fact that we are super transparent when it comes to showing our accuracy rates to our customers. This way, we hold ourselves up to the challenge of continuously improving the accuracy rates. Accuracy rates are one of our main differentiating factors especially because our focus is to save lives and prevent incidents.
Your company offers customer demos: What exactly do these show and how relevant are they to a typical wire and cable manufacturer?
Demos take about 60-90 minutes where we present our platform, how to set up rules and how it works in a real site example. We also provide examples of how we create value for the clients. It is very relevant to the wire and cable manufacturers as well. Most of the EHS use cases are addressing the typical high-risk activities in wire and cable sector.
How have employees responded to this usage? Is there a “big brother” sense of being watched that makes this a challenge for some potential deployments? Can this be used to look for lazy or slacking employees too?
This is a highly controversial issue that we take extremely seriously. At Intenseye we are proudly committed to 100% ethical AI. We are fully compliant with GDPR (i.e., faces are blurred, no facial recognition and biometrics are not captured). This is a red line for us! As a deeply woven company value, our team endorses “human performance principles,” i.e., We believe it is not the people who make violations willingly, it is the system/context around them which drives that behavior. Thus, our focus is to fix the systemic issues which make compliance difficult. We work with our clients very closely to ensure Intenseye’s real-time insights are being used to create a psychologically safe work environment where frontline workers are empowered to speak their minds and concerns with management because people are the best solutions to the issues. We’re thrilled to continue our journey to bring technology to the service of people to save more lives!
Is your company part of a vanguard that will take years and years to get traction?
We see ourselves as a company that is spearheading a bold movement with the noble purpose of saving lives. Most processes in EHS are still manually managed. People are using paperwork while they do their inspections only a few times a day. With AI technology, we automate this process and bring in real-time insights from the field. Also, we don’t think that this is going to take years and years. Our customers are very happy with Intenseye and many that have decided to scale up within a couple of months of the trial phase have already seen the value of it. I would like you to remember that technology is progressing rapidly, and if you want to use novel ways to save lives and prevent serious injuries at your workplaces and do this in a cost-efficient way (at a fraction of the overall annual PPE cost), I strongly recommend you to visit our website, follow us on LinkedIn and give us a shout!