AI video analytics for factories: safety, procedure and quality monitoring with SquintPRO

A Saudi plant usually has cameras for security and a supervisor on every shift, and the two rarely meet. Footage is watched only after something has gone wrong, while a person stepping into a robot cell or a skipped step at a workstation happens in seconds and leaves no record.

Computer vision in factories raises an alert when a condition you define occurs in a zone you draw. It works where the condition is visible; it does not see inside a machine or measure to a tolerance, which stays with your instruments.

For several sites, see technology for manufacturers with several plants; for permits and KPIs, our safety practice for plants and industrial sites.

Relatedtechnology for manufacturers with several plantssafety practice for plants and industrial sites

Hard hats, vests and coats are the strong case for PPE detection: large and worn where cameras see best. Gloves, glasses and ear protection need close views. Set grace areas where PPE is legitimately not yet on.

Around a robot cell, press or conveyor, SquintPRO alerts when a person enters the zone drawn around it. It does not replace guarding, interlocks or lockout; it adds an alert and a count of entries. Forklift zones in stores and loading bays work the same way; see forklift and pedestrian alerts on existing cameras.

Relatedforklift and pedestrian alerts on existing cameras

Many problems start with a step done differently from the standard operating procedure: a part fitted out of order, a check skipped, a tool left on the line. Where a step is visible from a fixed camera, SquintPRO can flag it. The condition comes from a trained library, a text description or a reference image of the right and wrong state, without code.

Start with short, visible, written steps; torque or temperature belong to the instrument. A team leader confirms or dismisses each alert, and the log shows where procedures slip by shift.

Line monitoring watches an area across a shift for missing or misplaced parts, and existing CCTV often suits it. Unit-level inspection checks each product at a fixed point, where camera position, lighting and line speed decide the result; see camera-based quality inspection on production lines. The quality release stays with your inspectors.

Relatedcamera-based quality inspection on production lines

An electrostatic discharge can damage a component invisibly, so ESD areas set rules on smocks, wrist straps and entry. A camera sees whether the person entering wears the smock; whether a strap is grounded is for the tester at the entrance.

In cleanrooms, gowning, doors held open and people out of hours are visible; particle counts stay with environmental monitoring.

SquintPRO is in live use in the Interkey group's own network-device assembly facility in Riyadh, run by Interkey Semiconductor. There it supports quality and operating discipline on the factory floor, including access and ESD compliance and safety governance.

Dust on domes, glare and night shifts lit for people move the result, so the pilot runs in them.

• A clear monitoring policy. Tell workers what is watched and why. SquintPRO does not use face recognition and does not identify people.
• Where footage is held. On the plant's own servers, or in the cloud for 3, 6 or 12 months.

Interkey's engineers carry out the survey, integration and support from Riyadh, in Arabic and English.

SquintPRO is licensed per camera stream, with unlimited use cases on a licensed stream, so adding an SOP check to a camera licensed for PPE adds no licence. A quote reflects the number of streams, the deployment model and any cameras, close-up stations or servers the survey finds are needed.