Oracle updates IoT Cloud, adds AI, VR, Digital Twin & Thread

Big Red has expanded its Oracle IoT Cloud offering, adding new AI and ML functions in the guise of Digital Twin and Digital Thread tools. New AR and VR applications were also on show, and Oracle is hoping that the new operation-wide visibility it can offer will gain the attention of industrial and enterprise customers – looking to embrace SaaS and PaaS offerings from the software and cloud provider.

Oracle is looking to snare any business looking to digitize part or all of its operations. Thanks to Oracle’s significant presence in the traditional business software and computing (on-premises and cloud) markets, it can count on those sales channels working in its favor – leading to conversations about its new IoT-augmented offerings, that it can pitch to CIOs looking for efficiency and margin improvements. If you’re a utility or grid operator, Oracle has you covered too.

As for the AI and ML features, Oracle is pushing the technology as the way to process data from those business applications, to provide anomaly-detecting analytics tools. These would help spot potential pain points in a supply chain, predict if a machine needs unscheduled maintenance to prevent costly downtime or emergency repairs, or be used to identify customers that might be likely to churn away from a service.

The overarching point of why AI and ML keep appearing in announcements like this is due to the quantity of data being generated. Simply put, there aren’t enough man-hours in the word to manually process each sensor reading from an IoT device, or a connected work machine – and automation only gets you so far, as you still need to turn that automatically harvested data into something that an operator or technician can understand.

It’s all well and good knowing that the core temperature of a water tank is 86C, or that a driveshaft is revolving at 16,000rpm, but without context, there is no understanding. You need to know whether those are within parameters, or if they have deviated from an expected norm – and the reasons behind those changes.

While a human engineer could monitor the live feeds from those machines, and spot if a problem was emerging, or an automated system could issue an alert once a threshold was exceeded, these systems are difficult to scale – either requiring lots more trained technicians to man the feeds, or a significant investment in an automated system that effectively just replaces those humans without actually improving the performance of the task at hand.

This is where the AI-based intuition comes in, acting as a system that can learn the intricacies of a particular process over time, to spot what is not normal behavior. Because it is an instance on a cloud computing platform, it can also be easily scaled up and down to match the number of processes that need monitoring – which promises to be much easier than using a human worker, and cheaper in the long-run.

In theory, the system can handle as much data as you can throw at it, and with more and more ‘things’ being connected in the IoT, these AI and ML processes look like the only way for industries to keep up with the exponential growth in data production and consumption.

In the unveiling, Group VP of Oracle IoT Cloud, Bhagat Nainani, argued that it is very hard for businesses to independently apply AI and ML functions to their operations – requiring expert programmers and data scientist. Oracle’s pitch is that it removes the need for those staff, with everything required built into the IoT Cloud and ready to be integrated with Oracles existing plethora of traditional business applications.

In the launch presentation, Oracle said that the IoT was at the top of a hype cycle, but that it still represented a $3tn market opportunity by 2020. Most notable was the estimate of 2 zettabytes of daily data production that is expected. This volume is why AI and ML technologies are critical for the growth of the IoT – hopefully altering the current paradigm, where very little of the data generated is actually used. This kind of tech provides the scale needed to process this volume, and do something useful with it.

The newly launched offerings are the Digital Twin for Supply Chain Management (SCM), the Digital Thread for SCM, and the integration of the AI and ML features. We’ve outlined the latter components above, and while some readers will be familiar with the Digital Twin concept, the Digital Thread concept is a little more obscure.

This thread is a method for connected data inside a supply chain that traditionally would be disconnected silos of information. When all of the data sources are brought together, inside one thread, it should give a business a much better view of the supply chain and manufacturing process.

But this requires both ways of getting that data into these cloud applications, and also a fairly substantial overhaul of business processes so that they can actually do something with this new data. The thread can be followed, to trace an item’s history, and Oracle is providing the software frameworks to let customers build these interconnected supply chains.

Digital Twins are virtual representations (models) of a physical object, which can be used in a number of applications. The Twin can be used to augment a CRM or ERP software application, and is essentially a way of digitally representing a physical object inside the digital world – bridging the IT and OT departments.

If a company designs the product, they can use the CAD drawings and schematics of the Twin in future designs, or to build accessories and housings for it. Those designs can also be used to create training and maintenance guides for technicians, or sent to customers and retailers as ways of adding the Twin to their own environments – essentially an additional service, on top of the product. In theory, testing will be much easier and more effective using the twin-system.

Oracle has also built solutions using a number of its applications, and these are where the AR and VR elements come in. In the demonstration, a worker was able to use a VR headset to train on new equipment before it was installed, which could help reduce the learning curve and potential costly mistakes as the worker gets into the swing of things, and can help troubleshoot problems once that worker has gone live in the field.

We saw something similar from PTC’s ThingWorx PaaS, when it integrated its purchase of Qualcomm’s Vuforia division. In both examples, we see technicians being guided through maintenance tasks by an AR application, who are also able to video-call for help when needed. Identifying codes (like QR codes) can be printed onto the devices, to aid quicker identification and localization. In both, the Digital Twin feeds into the AR applications.

Nainani said that Perfect Industry was the partner for the VR tech, supplying a software stack that was light enough to be used on devices like the Samsung Gear VR, rather than bulkier, tethered, and more expensive devices like the Oculus Rift. We remain skeptical that office-based managers will be clamoring to put on a headset and experience the overseas factory that is currently suffering an outage, but the in-field AR capabilities have clear advantages. He added that an unnamed German manufacturer specifically requested the training feature.

Oracle launched its first 4 IoT applications back in February, commencing with Asset Monitoring, Production Monitoring, Fleet Monitoring, and Connected Worker. Designed for specific business needs, they were offered in a completely managed SaaS package, which was built on top of the Oracle IoT Cloud. They supported extensive third-party integrations.

Notable customers for this first wave of products were Vinci (building automation, truck roll reductions), Mitsubishi Electric (factories, real-time production metrics), Noble Plastics (logistics overhaul, robotics integration), and SoftBank (electric scooter Mobility-as-a-Service project). The newly launched services, again built on top of the IoT Cloud, include Smart Connected Factory, Digital Field Service, and Digital Fleet Management.

In the past six months, Oracle says it has tripled its IoT application partner ecosystem, covering both device and system integration partners. This ecosystem currently comprises:

Advantech, Aricent, BAFX Products, Bosch,Cisco, Cisco Jasper, Dell, ELM Electronics, Estimote, Eurotech, Fathom, Freematics, Fujitsu, Gemalto, HMS, Honeywell, Huawei, Intel, Janztech,, LuvitRed, MatrikonOPC, MbientLab,  Mitsubishi Electric, Mobica, MultiTech, OSIsoft, Primal Sensors, Radius Networks, Rexroth (Bosch), Texas Instruments, Tridium, V2Com, Wind River (Intel), Xirgo Technologies, Zentri,

“Hitachi Consulting is constantly looking for industry leaders like Oracle to help clients and prospects harness the power of data and IoT to optimize operational and financial performance, out-pace their competition, and solve significant business problems,” said Garth Carther, VP North American Sales, Hitachi Consulting.

“Our longstanding strategic relationship with Oracle delivers IIoT solutions to digitize the physical word. In particular, Hitachi is leveraging Oracle IoT Applications – asset monitoring and production monitoring – that are uniquely engineered to meet specific industry and functional needs. These app have a depth of OT, IT, and IoT functionality that no competitor can match.”