The Internet of Things (IoT)—or, according to Cisco, the “Internet of Everything”—is starting to really heat up. We are finally seeing real use cases: automation in Australian mining, American farming, and the wineries of Spain; sensors with real-time monitoring; decision-making processing on driverless cars; and facial recognition software in phones and computers (queue the jokes about changing your face every ninety days to prevent intrusion).

There is a lot of confusion about IoT. Is it a network construct? Is it a distributed learning machine? Is it nothing more than a set of smart bulbs and a Google home? Personally, I feel it is quite simple. Microsoft’s term “the intelligent edge” sums up the area perfectly. It’s about sensors on vehicles that map their location; it is about facial recognition automatically logging you into your compute device, be it a smart phone, a Microsoft Surface, or your TV. Like all things under the sun, IoT is not a new construct—but it is an evolution. And it is an evolution whose time has arrived.

Let’s look at a little analogy in mining: an excavator extracts ore from the ground and transfers it to a large tipper truck, and that truck conveys it to an automated conveyor belt to the process plant.

The excavator and the tipper truck are run by humans, with all their frailties, including their capacities for making mistakes and getting injured. A mine is a very dangerous place. Removing humans from the equation would reduce injuries and mistakes, but until recently this has been almost impossible to do due to the immaturity of the network and the lack of computing power in the endpoint or the core systems. However, with the advent of 5G and ubiquitous high-speed broadband connectivity, we can now transfer a massive amount of data from active sensors and intelligent end devices to centralized, cloud-based analytic devices and systems that drive autonomous responses.

Currently, the mining industry is automating the last mile: the extraction of the ore from the ground. This is leading to drillers, excavators, and tipper trucks that are largely, but not fully, autonomous—ones that are managed remotely, very much like a drone. This is the halfway house we are looking at today, but it is still the Internet of Things.

In mining, this is being driven by machine manufacturers, like GE, Hyundai, and Hitachi, rather than by technology companies. These are operational companies, makers of large industrial machines. Both GE and Hitachi have been building up their expertise in this area. Hitachi and GE, with its Predix division, have gone one step further. They have created a new division by merging HDS, Hitachi Insight Group, and Pentaho into a new IoT-focused business called Vantara. This is an interesting development we will explore in a later article.

Dell, in what appears to be an attempt to remain relevant in this age of cloud computing, has made a serious move in the direction of the IoT, opening a new division headed up by Ray O’Farrell. It seems that Dell is positioning itself to become a distributed brain (not a traditional single brain model). It has announced a number of key projects:

  • Project Nautilus, which is software designed to ingest and query data streams collected from IoT gateways in real time;
  • Project Fire, which is a hyperconverged platform and part of the VMware Pulse family, with management, local compute, storage, and IoT applications such as real-time analytics;
  • Project IRIS, which is under development in RSA Labs and will be used to extend security analytics capability to provide threat visibility and monitoring at the edge; and
  • Project Worldwide Herd for performing analytics on geographically dispersed data.

Further, Dell will be investing in several IoT startups via its VC engine, Dell Technologies Capital. Edico Genome, FogHorn Systems, Graphcore, Moogsoft, and ZingBox were some of the early beneficiaries of capital.

It appears that there are now three pillars in IoT: the operational companies like GE, Hyundai, etc., which are adding sensors and AI to their equipment; the pure technology players Dell, Microsoft, and Cisco, which are attempting to build the smarts that will run over the hardware sensors; and finally Hitachi, which owns technology companies, and operational and hardware companies.

On paper, Hitachi and the operational equipment vendors are in a far better position to gain traction. This is not really a technology play, but rather a business play. When I purchase a new piece of equipment, I would rather receive it with equipment and software already integrated—a single throat to choke, so to speak—instead of having to deal with multiple disparate vendors. The intelligent edge is not just about strapping on a camera or other sensor; it is about mimicking human behavior when carrying out particular tasks.

IoT is heating up and intelligence is rising, but is it truly ready for mainstream? Well, yes. Even traditionally conservative areas like mining and winemaking are seriously looking at automation: not just auto-pickers or conveyor belts, but sensors to test the state of the water table to determine when vines must be watered or fertilized, and automating the mine head to remove ore, meaning being able to confirm exactly what’s being picked up by the excavators (it is no use just picking up any old piece of rock).

It is in this area that Dell wishes to play. Currently, IoT will send the sensor information to a centralised environment: either a public cloud instance or on-site corporate data centers. Dell is of the opinion that for real-time decisions, like an excavator needing to stop because it encounters something it doesn’t recognize, it is important for response to be instantaneous. Therefore, Dell thinks that there is a place for a local brain, a distributed brain that is nearer the edge to make instant decisions when a deer is in the headlights of your driverless car, or an undischarged explosive is in the ore rubble.

Who will win? I don’t know; it is really too early to tell. However, the intelligent edge has finally started to take shape.