Are alarms for businesses set to change forever? Nirovision, an Australian trailblazer in AI-powered computer vision believe this is to be the case. We interviewed their CEO, Jimmy Lee about their latest AI feature, Safety Agent.
Rapid: What exactly is Safety Agent?
Jimmy Lee: The best way I can describe it is, imagine employing a safety officer named Jack to monitor your security cameras all day. Jack’s role is to look out for safety issues and notify you if something requires attention. While Jack is already knowledgeable about what to look for, you provide him with additional information about your site and specific safety rules. For example, you inform Jack that hi-vis vests are required on the loading dock at all times.
Jack diligently monitors the cameras and alerts you if he observes anything of concern. For instance, Jack may notify you if he spots someone who appears to be injured, someone climbing a perimeter fence, or observes someone on the rooftop, dangerously close to the edge.
Now, let’s imagine Jack is not a human, but an AI Agent, that is not only knowledgeable, but also tireless, thorough and unbiased.
That’s the power of our latest breakthrough, Safety Agent – a world-first innovation that allows an AI agent to actively monitor your security cameras and promptly alert safety concerns.
Rapid: Wow! We know a lot of companies that could benefit from this. How can a workplace implement Safety Agent?
Jimmy Lee: First, we connect Safety Agent with your existing cameras. All you need is a dedicated server and you can choose which cameras Safety Agent should monitor. You can then optionally provide extra information about your site and any safety rules to help Safety Agent look out for the events that matter. That’s it. The Safety Agent will go to work and send descriptive email and/or phone notifications for specific issues so you can respond accordingly.
Rapid: What are some of the challenges with existing alarms?
Jimmy Lee: Nearly every business today has security cameras, and while the software to monitor camera footage has come a long way, the types of events you can be alerted about are extremely limited. For example, motion detection alarms can be triggered for all kinds of events, many of which end up being false alarms. Even person detection alarms can contribute to false positives. In general, traditional business alarms fail to answer questions like what caused the alarm, how do I know if it’s serious, who’s involved and what are they doing?
Typically, it is the responsibility of a security team, site manager or business owner to investigate and determine the answers to these questions. Safety Agent alarm solves these problems and offers additional benefits so you can get even more value from your existing camera infrastructure.
Rapid: Can you provide some examples of the types of things that Safety Agent can alert a workplace about?
Jimmy Lee: Yes. Safety Agent opens up a wide range of alarms that were previously not possible. This includes incidents that may serve as early indicators for more serious issues. Some examples include:
- Alert for someone lying on the ground
- Alert for someone on the roof near the edge
- Alert if a worker looks injured
- Alert for people physically fighting
- Alert for a person climbing a fence
Rapid: You mentioned Safety Agent can take in additional information about a site to ensure alarms are tailored to a workplace’s specific needs. How does this work?
Jimmy Lee: That’s right. While Safety Agent has been trained to look for basic safety issues, it also accepts extra information about your site and any safety rules that must be followed. This helps Safety Agent know what types of events and behaviours should trigger an alert. We provide a simple to fill text box where a workplace can add extra information to help Safety Agent.
Rapid: Can you tell us more about the descriptive alerts that Safety Agent sends?
Jimmy Lee: Safety Agent alerts include descriptive text about the situation so you can act swiftly. This is different to traditional alarms that do not contain any context or any content about what happened.
Conventional person alarms can’t detect nuanced human activity, such as appearing hurt or remaining still, or signs of potential injury.
Rapid: I think we’re all familiar with the issue of false alarms when it comes to monitoring. How does Safety Agent help in this regard?
Jimmy Lee: By analysing security footage using powerful AI capabilities, Safety Agent can quickly distinguish between non-events and genuine safety concerns. In the image below, is this person a security threat or is it just a cleaner working after hours? Safety Agent is smart enough to detect and not alert you for false positives.
Rapid: In addition to alarm notifications, what other insights does Safety Agent provide?
Jimmy Lee: Any events that trigger an alarm can be viewed in more detail. All these issues are logged so you can easily identify safety trends. This is important because incidents can serve as early indicators for more serious issues and surface repeat offenders. Being able to report on safety and compliance trends allows managers to better improve safety measures on site.
Rapid: Does Safety Agent replace security monitoring teams?
Jimmy Lee: No, Safety Agent empowers security teams by improving efficiencies and decision making, enabling a faster response to incidents. Security teams also benefit from having an audit trail of events that can be reviewed using the Nirovision applications and exported for auditing and record keeping purposes.
Rapid: Thanks for your time, Jimmy. The future of workplace safety is exciting.
Interested in Safety Agent for your workplace? Contact us today.