Predictive Maintenance
AggreGate as a preventive maintenance system enables the OEM equipment vendors and service providers to build their predictive maintenance (PM) solutions in a rapid and cost-effective manner. Allowing switching from a schedule-based or lifecycle-based to a "health-based" service, PM delivers one of the most important business optimization opportunities that was made available by the Industrial IoT technology.
Predictive maintenance technologies bring a univocal business value materialized by the higher equipment uptime, lower OPEX due to the reduced repair costs, and more strict Service Level Agreements (SLAs).
AggreGate ecosystem is adapted for ensuring the quick time-to-market for PM (preventive maintenance) initiatives. Here's how a typical project lifecycle looks like:
| Your equipment is fitted with various sensors to enable the telemetry acquisition. In most cases, the sensors are already there. |
| The "brain" of your unit (normally a PLC or an industrial PC) is extended by a small piece of software that is called AggreGate Agent. The agent takes care of bridging all collected data from the unit to a central server installed in a public or private cloud. |
| If modifying an existing unit firmware/software is not possible, an external Agent (a small PLC, e.g. based on the Tibbo Project System) can be added to the unit design. This external Agent is able to connect to the sensors directly or retrieve data from the main control computer. |
| The data gets reliably routed to the server and stored in a high-performance NoSQL database. In rapid deployment scenarios, it's also possible to import the historical data collected earlier. |
| Once a sufficient amount of data is available, it's necessary to teach the system which unit behavior patterns should be deemed negative. |
| Further operation continues in headless mode, AggreGate applies the machine learning and Big Data mining algorithms for predicting the unit health degradation. |
In many cases, predictive maintenance is not even an option, since the telemetry data received from a unit is a principal argument in solving the disputes between an equipment vendor and an operator. The Agent embedded into a unit can work autonomously, allowing the vendor to ensure that the temperature, humidity, inclination, acceleration, and other parameters were compliant during the unit transportation and storage.
Being deployed properly, the preventive maintenance software is predicting two valuable device health indicators: Time to Failure (TTF) and Remaining Useful Life (RUL). Those metrics are easily converted into the operating cycles, mileage, transaction counts, and other unit-specific metrics. The RUL estimation is a sophisticated matter, but it's proven to work quite good for the rotating machinery, such as fans, pumps and engines.
A predictive maintenance system is typically integrated with a Computerized Maintenance Management System (CMMS), automatically generating maintenance requests upon the equipment health degradation. Another advanced function is the service logistics.
Acting as a predictive maintenance software, AggreGate has deep knowledge about the equipment behavior. This knowledge can be used for the intelligent tuning and optimization of the device control logic and life cycle.
