IN Brief:
- Food manufacturers are expanding digital twin use for maintenance and anomaly detection.
- Amcor and Mars have both been cited as recent examples of multi-site deployment activity.
- Plants are linking analytics more closely with operations, alerting, and troubleshooting workflows.
Rockwell Automation has framed digital twins and industrial AI as practical tools for manufacturers moving from automation towards greater autonomy, and recent food-sector deployments suggest predictive maintenance is becoming one of the first areas where that shift is taking hold at scale.
Recent case studies point to food and packaging operators using digital twins to model asset behaviour, flag anomalies earlier, and connect monitoring more directly to plant decision-making. Amcor Flexibles, for example, has been using AVEVA’s MES, data services, and analytics tools to monitor around 200 blow and injection moulding assets across multiple plants, with the programme already identifying process issues linked to downtime and scrap. Early results pointed to a 2% potential reduction in unscheduled downtime across deployed sites.
Mars has also piloted a predictive maintenance programme for chocolate production across multiple plants using a digital twin environment running on Microsoft Azure IoT Edge. The aim has been to improve autonomous monitoring and give operators clearer visibility into normal and abnormal machine conditions before problems escalate.
The broader pattern is clear enough. Plants are no longer treating predictive maintenance as a standalone dashboard exercise. The focus is moving towards connected models, plant-level alerting, and repeatable deployment across multiple assets and facilities, which is where the gains in uptime and labour efficiency become more meaningful.
