My ARC Advisory Group colleague Sid Snitkin and I recently had a thought-provoking discussion on the future of the Industrial Internet of Things with Mark Morley, Director of Industry Marketing for Manufacturing at OpenText.
The Industrial Internet of Things offers the potential to enable industrial organizations to improve performance and enhance competitive ad-vantage — not only in an individual facility, but across company’s supply chain and throughout its value network.
Without a doubt, IIoT “things” – industrial smart devices that connect to the Internet and can collect useful data – will greatly outnumber people within a decade. Consider that IIoT things can include a company’s transportation assets, industrial equipment, products made, and even the containers that carry products across a supply chain.
OpenText, which is perhaps best known as a provider of enterprise content management (ECM) solutions that excel at managing unstructured data, is not a company that quickly comes to mind when thinking about potential suppliers of Industrial Internet of Things solutions.
When most people think IIoT, they think of using sensor data to improve business processes. Sensor data would usually be numeric data, and thus lends itself to being easily structured. Further, OpenText does not at this point, have any APIs to sensor devices.
However, as Mark pointed out, numbers by themselves may not mean anything until a rule set is built to interpret a particular data point. Once that rule set is built, either on the device itself or back in a corporate application, it can be acted upon. And OpenText does have a solution capable of interpreting quantitative data and kicking off the appropriate workflows.
While IIoT clearly has great potential, it is also clear is that the Industrial Internet of Things technologies are not fully mature, with new applications continuing to emerge. Through its acquisition of GXS, an EDI and B2B integration solutions provider, OpenText could potentially support an IIoT-powered order-to-cash process; something we have not heard about from any other supplier.
For example, the smart vending machine is one IIoT application that has been discussed in the media. Cantaloupe Systems, a San Francisco company that manages vending machines monitors a network of 100,000 snack and beverage machines scattered across the US. IIoT is being used to help employees monitor the vending machines from afar to assure that items remain in stock, sense temperature changes that would hurt product integrity, and even detecting thieves.
But IIoT vendor-managed inventory (VMI) processes are capable of more. At Cantaloupe Systems people are still part of the work flow. If an employee notices imminent product shortages, he or she must manually kick off the replenishment process. It would be possible to automate this by distributing the intelligence – moving intelligence from corporate systems and human beings to local machines. The IIoT sensors on the machine could have min/max replenishment logic and even kick off EDI purchase orders when the vending machine’s inventory drops below the preset minimum. When the machine is loaded with new inventory, it could kick off a receipt of goods electronic message, kicking off a payment process.
Beyond VMI, it is easy to see that IIoT could be used to enable a more automated order-to-cash processes across a variety of logistics, trading partner, and predictive maintenance scenarios.
In the smart vending machine example, the IIoT order-to-cash automation is saving dollars in the form of time and labor. But a faster order-to-cash cycle also improves a company’s cash flow position. The perfect order – the right goods, received on time, in the right condition – helps eliminate payment disputes and shortens the order-to-cash cycle for the supplier.
But consider the question of the vending machine receiving the right goods in the right condition. Vending machines could come prebuilt with a scanner attached to the inside of the vending machine, present for use when the machine is opened for stocking. The stocker could scan the SKU prior to putting it into the machine, and there would be proof that the right SKUs were being stocked. Similarly, the vending machine could have a digital camera, and the stocker could trigger a picture before slot-ting the product into the machine and thus proving the goods were not damaged.
Or, the scanning and pictures could be done using an automatic identification (AutoID) or smartphone device carried by the vending company stocker.
In short, there are choices about whether to drive the perfect order sensing functions to the vending machine or if it should reside on a device carried by the stocker. A study might be needed to determine which makes the most sense for a particular company and set of processes. And the answer that makes sense today, might not be the best answer in five years. IIoT also generates data. Lots of it. IIoT implementations often quickly become Big Data projects as well. But in many cases, not all of this data really needs to be stored. An alert is a notification that indicates a parameter – for example, the minimum threshold for replenishment – has moved outside of the desired range. Unless that minimum quantity level has been violated, and an alert is generated, does this data really need to be transferred to a corporate database and stored? Maybe, maybe not.
If not, the ability to distribute intelligence out to intelligent machines and devices can help prevent database bloat. Similarly, having some processes kicked off automatically at the device level could lead to a leaner and more streamlined application environment. But, inevitably, much IIoT data will need to be archived in corporate databases and integrated to enterprise applications. Again, companies will need to think through these questions carefully. There will be pros and cons to both approaches.
There were two insights I took from this discussion. First, IIoT is not fully mature and new applications for IIoT continue to emerge. Secondly, IIoT projects often become Big Data projects. Here, distributing more intelligence down to the device level could help prevent database bloat and streamline the corporate application environment. But these decisions are not clear cut. There are pros and cons associated with either distributing or centralizing both IIoT intelligence and data.
