Thingtrax Alternatives: Camera Agents, Vision Inspection or OEE Cameras

A ready meal line asks three camera questions at the end of the line. Did every tray get counted? Is every tray right? Does every label match the job that is running? Thingtrax sells an AI agent for each one.

A fourth question sits behind them: why did the line stop, and what did that cost? This page matches each job to the tools that lead with it.

A note on bias: Fabrico wrote this page and appears in it as an alternative. Hold us to the same evidence you would ask of Thingtrax. Fabrico comes first below because we wrote this page, not as a ranking. Other camera-based OEE tools are compared in the computer vision OEE comparison.

What Thingtrax sells (as of 11.11.2026, per its own site)

Thingtrax describes computer vision AI agents that observe production, reason in real time and trigger actions across food and beverage manufacturing (thingtrax.com).

• Counting Agent. Vision AI tracks every product in real time, with alerts when counts deviate from targets. It runs on a camera with an AI edge box and links to a PLC or MES through an API or a digital signal output (Counting Agent).
• Quality Inspection Agent. Inspects every product, detects defects and triggers automatic rejection. New SKU models are trained remotely, and the page lists links to MES or PLC for line stops, alarms and action triggers (Quality Inspection Agent).
• Label Verification Agent. Checks every label with vision AI and industrial cameras, including SKU, allergen declarations and date codes. Every inspected pack is logged, and API and CSV links connect to production plans, MES, ERP and quality systems (Label Verification Agent).
• Orchestration and Line Efficiency. A layer over the deployed agents that shows line-level OEE and bottlenecks, with one source of truth for downtime and performance data. The page says it works with existing MES, ERP and quality systems (Line Efficiency).
• Coming soon, per its homepage: custom agents, for listed use cases such as foreign object detection, seal or lid verification, and fill-level and portion checking (thingtrax.com).
• Company. The about page says Thingtrax began in 2015 at a Cisco and University College London accelerator, and names bakeries, meat processors and ready-meal producers. Thingtrax Limited is listed as active on the UK Companies House register (as of 11.11.2026).

The agent pages also show customer results. They are vendor-reported, so we leave them out.

Ask in the demo: how a stop gets its reason in the line efficiency layer, and where the counting camera would sit on a tray line.

When Thingtrax fits

Read its four product pages together and Thingtrax fits plants that:

• run food or beverage lines and want a camera check on every unit rather than a sample;
• need every label checked for allergens, date codes and SKU, with a digital record of each pack;
• want to count, inspect and verify with one vendor, starting with one agent and adding more across lines and sites;
• want the agents to act, by rejecting a product or stopping the line, and not only to report.

The alternatives below each lead with one of these jobs on their own sites.

Thingtrax's four jobs, and what else covers each one

Start from the job, then compare the tools that lead with it.

One camera can carry more than one job: Thingtrax's label page says plants can expand from label checks into quality inspection and OEE on one camera. Ask every vendor which jobs one camera position can do on your line.

Thingtrax and three alternatives at a glance

Every cell is drawn from the vendor's own pages on 11.11.2026; each section below carries the links.

See stops, reasons and OEE by line and shift on a demonstration line: book a Fabrico demo.

Fabrico: line time first, with a camera add-on where no signal reaches

This is our own product. Fabrico is a cloud OEE and light MES platform. Thingtrax's agents start from each unit they see; Fabrico starts from the line's clock: when the line stopped, for how long, the reason the operator gave, and what the lost time cost.

What runs today, per our OEE and MES overview and computer vision page:

• Machine signal first. A PLC, sensor or counter tells Fabrico when the line runs and how many units it makes. The operator taps in why it stopped, on a tablet or phone.
• Shift reports show OEE, downtime and losses in money by line and shift, on the web, iOS and Android.
• The camera add-on, on request, counts output where no usable signal exists, such as a hand pack-out table or an old flow wrapper. Click a moment on the shift timeline to open the camera add-on's video; in the standard setup it streams on request and is not stored in Fabrico.
• Your data leaves through a REST API and webhooks, or as Excel and CSV files.
• Still roadmap: a camera that flags a stop by itself, jam areas marked on its view, and a proposed reason that the operator confirms with one tap.
• Not Fabrico: product inspection, label verification and reject control. A dedicated system keeps those jobs, and its reject counter can be read into OEE quality.

Compared with Thingtrax: both can count with a camera. Thingtrax's counting agent sits beside its inspection and label agents; in Fabrico, a count from the camera add-on joins the PLC machines' data in one OEE view, next to the stop reasons operators enter.

Ask in the demo: what a camera count and a hand count from the same station look like side by side.

Fits plants that need the cost of downtime per line and shift before anything else, where PLC machines, old machines and hand stations share a line. See our OEE software guide and AI visual counting.

Cerrion: video AI agents across the whole line

Cerrion, based in Zurich, sells video AI agents that detect production issues and trigger a response in real time (cerrion.com). As of 11.11.2026, its food page lists ready-made detections for downtime, such as stuck cartons, product jams and fallen packs, and for quality, such as missing products in trays and damaged packaging. Its deployment page says the agents run on CCTV or industrial cameras already on the line.

Compared with Thingtrax: both sell camera agents that act on the line; Cerrion's pages lead with events such as jams and fallen packs, Thingtrax's with checks on each unit. Ask Cerrion how a detection reaches the PLC and the downtime report.

Fits plants that want cameras watching conveyors and transfers for jams and fallen product, per Cerrion's pages. Our Cerrion alternatives page covers it in full.

OAL: label and date code checks tied to the running job

OAL, based in Peterborough in the UK, sells compliance systems and robotics to food manufacturers (oalgroup.com). As of 11.11.2026, its site says:

• Its APRIL Eye AI vision system covers label verification, date code verification, PLU label presence, pallet integrity and seal integrity (APRIL Eye page).
• AI label verification, powered by Cognex, checks every pack at up to 200 packs per minute, including artwork, date codes, country of origin and promo labels.
• A camera on the line checks every date code, or an operator uses a hand scanner at the line terminal. The operator picks the job from the production plan, and wrong packaging stops the line for an investigation (label and packaging verification).
• The homepage says OAL helps food manufacturers achieve retailer compliance.
• OAL also sells printer control, robotic palletising, and OEE productivity software for food lines.

Compared with Thingtrax: both verify labels with AI vision for food plants. OAL ties each check to the job selected on its line terminal and sells printer control with it; Thingtrax's label agent shares a platform with its counting and quality agents.

Ask in the demo: how a new SKU's label gets into the product data.

Fits plants that supply UK retailers and want label checks, date code checks and printer control from one supplier, per OAL's pages. See label and code vision vs an OEE camera.

Which one fits your plant

• Thingtrax fits plants that want counting, inspection and label checks on one camera platform, with agents that reject or stop the line (its homepage).
• Cerrion fits plants that want cameras watching the whole line for jams, fallen packs and quality escapes (its food page).
• OAL fits plants whose first risk is a wrong label or date code on a UK retailer's product (its verification page).
• Fabrico fits plants whose first question is why the line stopped and what it cost, including hand stations with no machine signal (our computer vision page).

Running a label checker and an OEE tool side by side? Agree first which system owns the count of finished packs. For pack inspection beyond labels, see our Elementary alternatives.

Worked example: one tray line, two counters

This is made-up sample data for a chilled ready meal tray line; it is no customer's line and no product screen. The shift has 450 planned minutes and an ideal rate of 60 trays a minute: 27,000 trays at best.

The line was down for 54 minutes: 25 for a changeover, 14 for sealer film changes, 15 for labeller jams. That leaves 396 running minutes, so availability is 88.0%.

At the ideal rate, 396 minutes is worth 23,760 trays. The camera above the hand pack-out saw 21,344 good trays leave, and the labeller's reject counter logged 436 more, so the line made 21,780 trays: performance is 91.7%.

Quality is 21,344 of 21,780, or 98.0%. Multiply the three: 88.0% x 91.7% x 98.0% = 79.1%, and 21,344 of 27,000 possible is also 79.1%.

Now the second counter. The depositor's PLC counted 22,320 trays, 540 more than the 21,780 the line made. Those trays left the line in between, for example pulled by hand at the sealer, and nothing logged them. On the depositor count, OEE would read 81.1%, which flatters the line.

So count finished product where it leaves the line, and check that count against a hand count before you trust it. See how accurate camera OEE is and how many cameras a line needs.

Questions to put to every vendor on this list

Use this list in every demo, ours included.

• Which of the four jobs is it for? Count, inspect, verify labels or measure line time. Ask which run on a line like yours today.
• What happens at a changeover? Who sets up a new SKU, label or tray, and how long the check is off while that happens.
• How does a stop get its reason? From the operator, from a fault code or from the camera, and what happens to the stops nobody explains.
• Which count wins? If the PLC and the camera disagree, which one feeds OEE and your ERP.
• How long is the pilot, and what is the yardstick? One line, a set number of weeks, and a hand count or QA log as the reference.
• What goes into the quote? Lines, stations, cameras and edge boxes, brackets, network and training. Send every vendor the same station list.

For the OEE half of a trial, follow our computer vision OEE pilot plan. For food lines, see computer vision for food manufacturing and computer vision for ready meal production, and for the wider picture our FMCG buyer's guide.

Cameras over people: agree it before they go up

Cameras at pack-out see hands. Cerrion, for one, says on its security page that it anonymises people in detection clips on its on-site server; ask each vendor what its cameras record and who can see it.

Fabrico's camera add-on is aimed at the product flow and, in the standard setup, streams video for a stop only when someone opens it. Agree camera positions with your works council first; see GDPR and line cameras and cameras without surveillance.

What people ask next

Is Thingtrax OEE software?

Its line efficiency page describes line-level OEE, bottleneck detection and one source of truth for downtime and performance data, built on its AI agents. To compare it with OEE software, run both on one line and check how stops get their reasons.

Which Thingtrax agents can you buy today?

Its homepage names three out-of-the-box agents, for counting, quality inspection and label verification, plus the line efficiency layer. Custom agents are listed as coming soon, with use cases such as seal or lid checks.

Does Thingtrax connect to a PLC or MES?

Its counting page says the agent integrates with a PLC or MES through an API or a digital signal output, and its homepage lists ERP, quality, maintenance and MES systems. Ask Thingtrax to confirm the link for your PLC model.

How is Fabrico different from Thingtrax?

Thingtrax's site describes camera agents that count, inspect and verify labels. Fabrico is built around line time: the machine signal gives stops and counts, the operator gives reasons, and reports price the losses by line and shift. A camera add-on counts where no signal reaches. Fabrico does not inspect products or check labels.

Can a label checker's rejects count in OEE?

They can, when the label system puts out a reject signal or keeps a reject counter. The OEE side then books those packs as quality losses, and the reason each label failed stays in the label system.

Want stops, reasons and camera add-on counts shown on a line like yours? Book a Fabrico demo.