Last breakout session at the IBM Big Data and Analytics day in on the Internet of Things and Predictive Asset Maintenance. The IoT is creating new opportunities in all sorts of industries from route optimization to border control, power management in buildings to managing ATM infrastructure. Increasingly organizations want to respond in real-time to things being detected by sensors. The IoT helps with the Sense and Build elements of the Sense-Build-Decide-Act cycle, helping you establish that something has happened and the context of what happened. As more IoT devices have actuators as well as sensors it also enables automated Act once a decision is made.
For IoT IBM is leveraging its Bluemix PaaS offerings – both IBM components and partner components like geolocation services. They have created an IoT starter kit with Informix for data capture, data connectors, and Node Red for stringing together IoT devices with Javascript. The kit allows companies to connect devices using an MQTT based open standard, collect and manage data about the devices, assemble these elements using Node Red Javascript and then build applications on Bluemix that leverage all of this.
IBM talks about these solutions involving
- A device tier
- An edge tier gateways for aggregation and some processing
- A cloud tier for applications
IBM is positioning their Informix database as one that can work in all three tiers, providing a common data infrastructure, but in particular it can act as a time series data store in the edge tier. Partners, such as those building smart home gateways, are embedding Informix to do capture data over time as it streams in from connected devices. Some processing can then be done on this gateway while data can also be pushed up to the cloud for more advanced analytics (though there is no current way to push the resulting analytics or rules based in to the gateway).
IBM has been integrating their Predictive Maintenance and Quality solution with this IoT platform. The Predictive Maintenance and Quality solution is designed to improve reliability, customer service certainty, maintenance efficiency and total cost of ownership. It collects and integrates data from connected devices, generates predictive analytic models from this data, adds these insights to the data to drive recommended actions using the decision management engine that can then be taken automatically or recommended to a human user.
The new version has improved support for quality metrics, better integration with Maximo, new APIs and improved maintenance modeling. Plus some specific content around the mining industry. Many different data streams about assets (MTBF data, maintenance record text, operating conditions data and models of wear/operations) are combined and used both to provide dashboards and reporting and to drive automated decisions to recommend actions. Some of this is at the SKU or product level (how are our Electric Vehicle batteries working overall) while others are at the specific instance of a product or asset (advising a specific driver about the behavior of their specific battery).
I believe that the value of IoT is only going to be realized when analytic decisions are embedded in the systems that are connected to the devices. This combination of Predictive Maintenance and Quality with IoT is a good example.
Comments on this entry are closed.