IntelliAM AI has secured a further purchase order from a major agricultural processing and commodities business, taking orders from the unnamed customer above £450,000 over the past twelve months and extending its industrial artificial-intelligence platform into an EMEA trial.
The project will test IntelliAM’s machine learning and AI capability within the customer’s regional operations. The company says the work could lead to broader deployment of the first two layers of its platform: predictive maintenance and reliability intelligence, followed by operational intelligence connecting equipment condition with wider production data.
Neither the value of the latest individual order nor the sites covered by the trial have been disclosed. The announcement therefore represents a further commercial step with an existing customer rather than a confirmed large-scale rollout, with expansion dependent on the outcome of the trial.
IntelliAM’s reliability layer combines information from industrial sources including sensors, programmable logic controllers, computerised maintenance-management records, asset metadata, manuals, and maintenance schedules. The operational layer adds production and process information so that equipment condition can be considered alongside throughput, settings, quality, and operating state.
That context is central to predictive maintenance. Detecting an unusual vibration or temperature is comparatively straightforward once instrumentation is installed; determining whether it represents a developing failure, a speed change, a new product, or normal process variation is more difficult.
A useful system therefore has to connect condition signals with information about what the machine was doing when the change occurred. Without that connection, maintenance teams can receive large numbers of alerts that require investigation but do not lead to useful intervention.
Industrial users are increasingly moving predictive-maintenance programmes beyond isolated monitoring pilots. Yorkshire Water’s extension of electrical-signature monitoring, for example, combines continuous asset data with specialist review and defined corrective actions rather than treating analytics as a stand-alone dashboard.
IntelliAM is using an existing customer relationship as the route into the broader operational layer. The company followed a similar expansion model with Mars UK earlier in 2026, when purchase orders worth £425,000 extended deployment across six sites covering chocolate, petcare, and Wrigley operations.
Its stated approach is to begin with asset and reliability information before increasing the scope of analysis as customer systems and operating practices are connected. That sequencing gives maintenance teams an initial use case with measurable engineering outcomes before more production variables are added.
IntelliAM reported group revenue of £5.26 million for the year to 31 March 2026, up 64%, while annual recurring revenue reached £1.65 million. The £450,000-plus value of orders from the agricultural-processing customer is therefore material relative to the present scale of the company, even though the latest EMEA project remains a trial.
Agricultural processing provides a demanding operating environment for predictive analytics. Production assets can run for extended periods, handle variable raw materials, and combine conveying, separation, thermal processes, storage, and packaging. Changes in feedstock or operating recipe can alter machine behaviour without indicating a fault.
Models trained without sufficient operational context can therefore mistake legitimate process variation for equipment deterioration. A system that links machine condition with production state has a better basis for distinguishing an emerging mechanical problem from a change caused by normal operation.
Integration with existing factory systems remains another constraint. Manufacturers rarely operate a uniform estate of new equipment, so analytics platforms have to work across mixed sensors, controls, maintenance databases, and manually maintained engineering records accumulated over many years.
IntelliAM says its reliability layer is designed for that mixed environment and includes human validation of work orders, keeping automated analysis connected to engineering judgement rather than allowing algorithms to issue maintenance actions without review.
The challenge increases as the platform moves from reliability into operational intelligence. Maintenance models can be assessed against downtime, failure avoidance, and work-order quality; production optimisation introduces additional trade-offs involving throughput, energy, quality, waste, and scheduling.
The EMEA trial will therefore test more than whether IntelliAM can detect equipment anomalies. It will show whether data gathered for asset health can be combined with production information without creating conflicting recommendations between maintenance and operations teams.
IntelliAM has already demonstrated that it can expand within established customer accounts. The next commercial evidence will be whether this trial converts into recurring deployment across additional sites or business units, and whether both reliability and operational intelligence become part of the customer’s routine production systems.
