Database giant Oracle joined embedded software delivery specialist Wind River to present a session at Internet of Manufacturing 2017, held in Munich this February by Internet of Business. Oracle’s Lionel Chocron and Wind River’s Emaka Nwafor fired up a joint analysis to explore the world of IoT-empowered machinery control.
Devices that can not fail
Wind River has been in business for 30 years and has always had a primary focus on the embedded software market, even before we started using the Internet of Things (IoT) as a piece of terminology.
With specialisms in aerospace and defense as well as the industrial market, Intel acquired Wind River back in 2009. The firm says it works to build devices and machines that ‘just can’t fail’ in terms of their resiliency. NASA’s Mars Rover runs Wind River – and its next service is a long way off.
Wind River asserts that the IoT can enable new operational business models:
- Maintenance. Both predictive and prescriptive maintenance are now possible. (Previously, companies only requested maintenance when machinery was broken or when maintenance was a formally scheduled event.)
- Consumption models. Companies can now move to ownership on an as-a-service basis, not just for technology, but for a whole range of business assets.
- Business analytics. There is a move from static analysis to real-time big data analytics.
- Service models. Companies are now in a place where service has moved from a centralized service function, to more self-service, self-guided business service models.
Oracle’s Lionel Chocron explained the conceptualization behind his team’s approach to an IoT-enabled modern supply chain model. The result is Oracle Digital Field Service (this is Wind River technology integrated into Oracle technology), which embodies the progression in service provision from being reactive to proactive.
Related: IFS enters era of ‘maintenance by information’, buys Mxi Technologies
Pump up the servicing
“Trips to customers – to fix a pump, or a motor, or whatever – are done on a reactive basis by field service engineers and, crucially, these jobs are typically only fully fixed perhaps three out of four times,” said Chocron.
The field service engineer can now use a dashboard to view the operational status of the assets (pumps and other equipment), to study everything down to the amount of pressure that is being exerted on the bearings in the machine itself and produce stats for:
- OTA (over the air testing)
- Dispatch times
Using all these stats, they can build up an actual percentage figure of failure probability, time to failure and an ‘up’ or ‘down’ arrow to denote performance trends – in other words, is performance deteriorating or still good?
They can also, in some instances, upgrade software on these types of machines and perform diagnostics remotely.
Related: Predictive maintenance and the future of manufacturing with IoT
Wind River’s Emaka Nwafor explained that not everything here is easy yet. For example, there is a high cost associated with integrating applications with edge devices. There is also (currently) a lack of trust in IoT systems from end-to-end (the passage of data through the whole system brings accuracy into question) and so companies are not prepared to analyze the data that is generated and captured.
Wind River uses Oracle Asset Management applications and provides other layers of technology that essentially exist as pre-built applications. There are other challenges in terms of device lifecycle management, managing change and managing scale. An initially deployed system may perform well during its POC (proof of concept) stage, but complexity in terms of full-blown deployment (with real data) makes things a lot tougher.
Related: Microsoft adds predictive maintenance to Azure IoT suite
Our push-button future
In the future, the near future perhaps, when all this stuff really works, companies will enjoy push-button device integration, due to the existence of pre-deployed device agents (the code that sits on the actual devices), while also benefiting from pre-built enterprise integration, so that devices can be engineered into the back-end processes and existing applications.
We’re not quite there yet, but we do know where we are going. Just a couple more field service engineer visits should get us there. Cup of tea, anyone?