Largely, conversations round linked staff start with wearables like good helmets, good watch, panic buttons, and biometric bracelets. That is the place most press protection and vendor exercise has occurred. Whereas the trade has been constructing the ‘linked employee’ idea round new {hardware}, it has ignored a sensor that has already been mounted on almost each wall of the jobsite: the CCTV digicam.
The primary goal of putting in CCTV cameras is safety surveillance and never employee connectivity, so they don’t typically get talked about in conversations on linked staff. I see this as a class mistake, not a expertise one. The expertise is already current however what’s lacking is the layer of reasoning that turns a passive video feed right into a stay security sign.
Context, Not Simply Vitals
Wearables are efficient at reporting bodily capabilities. A wristband, as an illustration, can flag a employee’s irregular coronary heart charge or an surprising fall, however it can’t clarify the rationale behind it. A employee’s coronary heart charge could spike from warmth, exertion, or as a result of a reversing excavator is closing in behind him. Both means, a biometric sensor alone can’t present the precise explanation for the change.
That is the place imaginative and prescient AI (synthetic intelligence) brings one thing essentially new and useful. Digicam-based monitoring doesn’t simply detect our bodies in movement. It may possibly additionally present precise context, like affirm PPE (private protecting tools) compliance, or flag when somebody enters a restricted space. This situational context is what no wearable can infer.
Each approaches are complementary, not contradictory to one another. The worth is in how the 2 indicators join. When a wearable flag an irregular coronary heart charge in a confined area, that alert alone tells a supervisor little past a quantity. If the identical sign triggers a test in opposition to the digicam analytics already working on that zone, confirming occupancy time, proximity dangers, and whether or not air flow tools is lively, the alert stops being a uncooked biometric spike and turns into a selected one: a named location, an identifiable hazard, and a timestamp a supervisor can act on.
The wearable initiates the alert, the digicam provides the context, and solely the mixed output reaches the supervisor. That sequencing, not the person sensors, is what makes the system linked.
I consider the expertise to make jobsites safer already exists normally. What’s lacking is the self-discipline to attach what’s already there earlier than shopping for one thing new.
A wearable system acknowledges the precise capabilities of the physique, whereas camera-based AI is helpful in relation to the exterior features of the work carried out. Put collectively, each complement a linked employee programme that captures each inside and exterior sign, as a substitute of treating both one as the total image.
Nevertheless, all this doesn’t make imaginative and prescient AI a substitute for wearables and sensors. Mud, poor lighting, occlusion, digicam placement, and the like are some actual limitations of imaginative and prescient AI techniques on a working web site. These techniques require deliberate configuration, coaching, and validation fairly than assuming they are going to work with none further efforts.
My level shouldn’t be that the digicam replaces the wearable. It’s that the majority linked employee methods begin with a {hardware} buy earlier than they ask what the positioning already has. A digicam already mounted on a wall, feeding right into a system that may purpose about what it sees, shouldn’t be a brand new sensor to purchase. It’s an present one which has been underused. Earlier than including one other system to a employee’s physique, it’s price asking what the infrastructure already on the wall may inform, if anybody had constructed the layer to hearken to it.
Concerning the Creator: Gary Ng, CEO and cofounder of viAct, has spent greater than 10 years advancing digital transformation in development. Underneath his management, viAct’s AI‑pushed laptop‑imaginative and prescient expertise is reshaping how jobsites are monitored—decreasing accidents, stopping delays, and boosting productiveness and margins. He additionally contributes to the following technology of trade expertise as a visiting school skilled at The Hong Kong Polytechnic College. HE might be reached at gary@viact.ai