Predictive Maintenance: How a Sensor-to-Cloud Retrofit Drives Benefits

Summary

If you're a machine builder or user maintaining hundreds of machines, then failures are the things you dread most. You know you need Predictive Maintenance, but the cost is often prohibitive. This article discusses a retrofit approach that uses an edge device to continuously measure a machine's critical parameters and send the sensor data to the cloud. Configuration is automatic and the solution bundles a ready-made cloud subscription to let you get on-line in minutes. An API enables data from machines located anywhere in the world to be monitored centrally using your choice of commercial off the shelf software.

The challenge

Machines and processes that have been working for many years are often difficult or impossible to monitor. Failures can be frequent and breakdowns costly. For users and machine vendors with a widely distributed installed base, the maintenance challenge can be difficult. What's needed is a way to oversee everything from a central location to facilitate predictive maintenance strategies remotely. Plenty of packages exist to do this but retrofitting them can be intrusive and expensive.

Predictive maintenance is a type of condition-based maintenance that monitors assets, such as equipment and machines, using sensor devices. These sensors supply real-time data that can be used to predict when the asset will require maintenance.

For OEMs and end users, the financial benefits of predictive maintenance monitoring of their machines can be dramatic. If failing equipment can be repaired or replaced in a timely manner, expensive downtime or even catastrophic failure can be avoided.

There is a huge base of installed machines in the world, typically referred to as “brownfield." This represents a tremendous opportunity for a predictive maintenance solution that can be “bolted-on” existing equipment to provide condition-based monitoring.

For this solution to succeed, it must include: Support for many machines in a single eco-system; simple, quick, and remote commissioning; flexible choice of attached sensor devices; built-in security. 

Predictive maintenance benefits

Assets are costly and downtime is expensive. For OEMs and end users, the financial benefits of Predictive Maintenance can be dramatic, including: 10X Return on Investment (ROI); 25-30% reduction in maintenance costs; and 35-45% reduction in downtime.

These benefits accrue in a variety of ways. For example, a machine builder can retrofit machines in the field with sensors and a sensor “edge device." The sensors may be accelerometers to monitor vibration, or thermocouples to monitor machine temperature. This data has value—a change in vibration or temperature can indicate a pending bearing failure or motor failure. The sensor edge device collects the data from the monitoring sensors and transmits it to an IT or cloud-based application for display and analysis.

Cloud-based predictive maintenance solutions can operate in isolation from a machine’s existing control system and therefore have no impact on existing automation. Their primary goal is to collect and aggregate data from equipment-based sensors that are installed anywhere in the world. This data can be used to identify upcoming problems. Failing equipment can be repaired or replaced in a timely manner.

Enormous potential with Brownfield Retrofits

Where hundreds of individual machines are installed, perhaps globally, the potential is huge. Cloud-based systems make it easy to monitor machines located anywhere without leaving the office. 24/7 monitoring makes it very easy to learn operating profiles.

Application lies in almost every industrial market, including the following:

• Machine tools
• Material conveyors
• Assembly machines
• Handling systems
• Pick-and-place

Retrofitting machines in the field with sensors and a sensor “edge device” is key to building a low-cost predictive maintenance system that is both secure and simple. The sensor edge device aggregates data from a wide choice of sensors, handles local communication processing, and provides an Internet uplink to the cloud. Once in the cloud, data can be fetched by a customer application for additional analysis and processing.

Ready to go, real-time monitoring

Cloud connectivity is bundled with sensorEDGE, and automatically invoked once a user account has been initiated. Users lease the cloud services, with ongoing costs as little as a few dollars per month.

sensorEDGE is automatically set-up in minutes. Operating software is downloaded, and the connected sensors are scanned and configured remotely. Then, sensorEDGE starts transmitting data to the cloud at the rate of 1 measurement/second.

A browser is used for device management. The user can watch the uploaded data from any location in real-time via a dashboard that displays each data stream as a widget. Line, Gauge and Bar Chart widgets can be selected. This dashboard can act as a simple real-time monitor, allowing machine responses to be watched from any location.

The sensorEDGE package includes the Rest API, a secure cloud doorway that allows third-party software to stream data from the cloud. In this way, data from hundreds of machines can be centrally monitored using relevant commercial off the shelf software.

The sensorEDGE solution represents a low-cost and easy-to-implement way to get data from sensors retrofitted to a machine, into the cloud for monitoring worldwide. Machine manufacturers could develop a “Retrofit Kit” that includes the sensorEDGE and appropriate sensors for the application.