AWS Partner Network (APN) Blog

Predictive Maintenance at Scale with Treon Flow and AWS IoT Core

By: Parascovia Digori, Partner Solutions Architect – AWS
By: Fio Piriz Sapio, Partner Solutions Architect – AWS
By: Mikko Nurmimaki, Product Marketing Director – Treon

Treon logo
Treon Connect with Treon

Unplanned equipment downtime costs manufacturers tens of thousands of dollars per hour (Senseye-Siemens 2022 True Cost of Downtime report). Amazon Monitron made machine learning-based condition monitoring straightforward to deploy, and many industrial teams rely on it to detect equipment issues before they cause downtime. Treon Flow, which incorporates core Amazon Monitron technology, gives these teams a natural continuation path as they extend their programs with artificial intelligence (AI)-powered analytics and maintenance workflows.

This post focuses on three things: the industry shifts shaping predictive maintenance, the AWS services that let Treon Flow scale to thousands of monitored assets, and an example from a food and beverage distribution deployment. For teams that want to bring existing Amazon Monitron data together with Treon sensors and other systems, the Guidance for Integrating Amazon Monitron with Treon Connect on AWS provides a reference architecture.

Where predictive maintenance is heading

Three shifts are reshaping how manufacturers approach asset reliability:

Contextual analytics. Traditional vibration monitoring tells you that something changed. Contextual analytics combines vibration, ambient, and operational data to tell you why it changed and what to do next. This moves maintenance teams from reactive investigation to guided decision-making.

Automated maintenance orchestration. As machine learning models mature, maintenance workflows become increasingly autonomous. Rather than generating an alert and waiting for a human to triage, automated systems analyze the root cause and evaluate severity. They then recommend actions, assign work orders, and verify repairs. All of this operates within defined guardrails set by reliability engineers.

Technician productivity tools. Frontline technicians spend significant time searching for repair procedures, interpreting sensor data, and documenting work. AI-powered tools in mobile applications bring relevant procedures and sensor context to technicians, which reduces search time and speeds up repairs.

Treon is building these capabilities into the Treon Flow roadmap on AWS IoT Core, Amazon Elastic Kubernetes Service (Amazon EKS), and Amazon Simple Storage Service (Amazon S3). Together, these services provide the device connectivity, compute, and storage that this architecture needs at industrial scale.

How Treon Flow extends Amazon Monitron technology on AWS

Many manufacturers have already invested in predictive maintenance: sensors deployed, baselines calibrated, and teams trained. With Treon Flow, teams can continue with a familiar sensor design, data formats, and existing ways of working while adding new capabilities. Treon Flow runs on AWS IoT Core, which is built to handle high-volume Message Queuing Telemetry Transport (MQTT) sensor traffic at global scale. Amazon EKS scales machine learning inference workloads as asset fleets grow, and Amazon S3 provides durable storage of sensor history and model training data. This AWS foundation is what allows Treon Flow to scale from a single production line to thousands of monitored assets without re-architecting.

For teams currently using Amazon Monitron, extending their program starts with what is already in place. Treon’s Industrial Node C sensor is based on Amazon Monitron technology, which means familiar hardware, data formats, and operating principles carry forward. Treon Flow then adds capabilities that extend these condition monitoring programs:

  • Broader anomaly detection. Additional key performance indicators (KPIs), alongside vibration root mean square (RMS) and temperature, including acceleration RMS (aRMS), help detect failure modes such as belt misalignment.
  • Customizable AI model training. Operators optimize anomaly detection self-learning periods for their specific asset types and operating conditions, improving detection accuracy over time.
  • Complete maintenance workflows. Treon Flow manages the entire maintenance process in one place: event and fault information, a visual site map with asset health, work assignment, repair guidance, and completion reporting, with a mobile-first experience for field technicians.
  • Flexible site management. Configurable hierarchies and visual layout maps show site-wide asset health in one view, scaling from a single production line to thousands of assets globally.
  • High-precision sensing for critical assets. Treon Industrial Node X provides higher measurement accuracy and broader KPI coverage for critical assets, so organizations can monitor different asset types in a single solution.

These capabilities let manufacturers keep their current predictive maintenance practices running while gradually adopting more automated and prescriptive workflows. Figure 1 shows Treon Industrial Node C, a sensor based on Amazon Monitron technology.

Treon Industrial Node C wireless vibration sensor shown with the Treon and AWS logos.

Figure 1. Treon Industrial Node C keeps the sensor design Amazon Monitron teams already know, while Treon Flow adds analytics and maintenance workflows on AWS IoT Core.

Predictive maintenance in practice

A large food and beverage distributor faced repeated gearbox failures on a critical dispatch conveyor line. These failures delayed orders and had a significant financial impact through lost throughput and retailer penalties.

The distributor deployed Treon Flow with wireless sensors. Using nearly continuous monitoring and machine learning analytics, Treon Flow detected an early bearing defect several days before failure, which prevented a production stoppage on the outbound conveyor line. The deployment also improved delivery reliability, fulfillment performance, and operational efficiency.

Conclusion

For manufacturers who have invested in Amazon Monitron and the practices around it, Treon Flow offers a way to keep building on that investment. It incorporates Amazon Monitron technology and runs on AWS, so teams can build on a familiar sensor design, baselines, and maintenance processes as their programs grow.

Treon’s roadmap for Treon Flow includes contextual analytics, automated workflow orchestration, and technician productivity tools. Organizations that start today will already have their sensors connected, baselines established, and teams trained on a system that evolves with them. To get started, review the Treon Flow offer on AWS Marketplace. Amazon Monitron users can contact Treon to explore extending their programs.

To learn more about predictive maintenance with AWS Partners, read Building a Predictive Maintenance Solution Using AWS AutoML and No-Code Tools and Real-time Data Monitoring with AWS IoT Core and Imply Data on the APN Blog.

Connect with Treon


Treon – AWS Partner Spotlight

Treon is an AWS Advanced Tier Partner specializing in AI-powered predictive maintenance for industrial operations. The company supports more than 200 customers worldwide across manufacturing, material handling, and logistics sectors. Treon Flow has completed the AWS Foundational Technical Review (FTR).

Contact Treon | Partner Overview | AWS Marketplace