OBSIDIAR

A machine cannot be told to be careful.

Industrial perception for the factory floor. Three products, three jobs, three levels of authority.

Scroll
92%
of people our cameras caught had no light-curtain trip — blind sensors never saw them
4
Sensor modalities — eyes, heat, presence, data
27%
of machine-stop time was needless — uptime our AI recovers
24/7
Detection in dark, smoke, fog and hazardous zones

Three Jobs. Three Products.

One platform cannot be a protective device and a quality inspector at the same time without the failures of one contaminating the trust in the other. So they are separate products, with separate authority, deployed separately — and that separation is deliberate.

It Comes Down to One Verb.

The rule we design to internally, and the fastest way to route a requirement to the right platform.

THE CELL

If the risk is a person inside the working volume — walking into a robot cell, still inside when someone hits restart — it is Monolith. It guards the space around the machine and the conditions under which the machine may run at all.

THE TASK

If the risk is at the point of operation — a hand reaching into the blade, the press, the cutting head — it is Forge. It stops the machine on a breach, and judges PPE, posture and task sequence at the same time.

THE LOAD

If it measures, counts or grades the product, it is TesseRack. It raises a quality flag or a reject so a fault is corrected upstream, not discovered at the customer — and it is the one platform here that never stops a machine.

Two protective products, one quality product — and why that line matters

Monolith and Forge are both protective: each can bring a machine to a stop, and each is scoped, tested and defended as a safety function in its own right. They are separate products because they answer different questions — Monolith asks who is in the cell, Forge asks what is at the tool — and a sensor placed to answer one of those well is rarely placed to answer the other.

TesseRack is the one that never stops anything. A quality reject and a safety stop must never be read as the same event: if a pallet-quality flag could look like a guarding failure, every nuisance reject would erode confidence in the thing that actually protects people — and an operator who stops trusting the guarding starts bypassing it. Bypassed guarding is the single most common root cause behind guarding-related fatalities.

FMCG and packaging · automotive and components · tyre manufacturing · renewables and wind · warehousing and logistics.

Every guard on a factory floor is a compromise between how safe the cell is and how fast it runs. We build for the sites that refuse to choose.

Obsidiar

Protect Your Site.
Deploy Obsidiar.

Tell us about your facility. We'll identify which platform fits the problem, size the sensor configuration, generate a full BOM, and walk you through the compliance picture — at no cost, no obligation.

Response Time
Within 24 hours
Assessment Cost
No charge
Coverage
Pan-India
Deployment
4–12 weeks from order
WhatsApp