“Digital technologies are transforming industry,” says the UK Government-commissioned Made Smarter report. “Emerging technology breakthroughs in fields such as AI (artificial intelligence), robotics, and the Internet of Things are significant in their own right. However, it is the convergence of these IDTs (industrial digital technologies) that really turbocharges their impact.”

The Industrial Internet of Things (IIoT) cuts out the middlehuman, and lets devices communicate directly, sharing data as easily as toddlers share germs. Every day, 5m devices link to each other, and by 2020 there will be 20bn data-sharing objects in the world. That’s the forecast of the Made Smarter report, which concluded that investing in Industry 4.0 could boost manufacturing growth by 1.5% to 3% every year.

The IIoT promises improvements at every level, from production scheduling to customer satisfaction. With a cleaner, safer workplace and less unplanned downtime, small gains in efficiency swiftly add up to big savings.

Companies like Sight Machine, which originated in Michigan’s automotive industry, now offer data analytics to enterprises of all shapes and sizes. By bringing together diverse types of data into one system, managers can oversee the whole manufacturing and distribution process on one dashboard. In real time, the information flows in and the picture updates.

Digital design and additive manufacturing techniques disrupt our idea of how work flows through a factory. If parts can be custom-made within days or hours, every job becomes a bespoke job. Luckily, supply chain management is now delegated to powerful computers that can handle the new complexity.

Thankfully, AI anticipates problems before they happen. Whether it’s a machine about to break down, an understocked warehouse, or an over-tired worker, your Machine Learning algorithm will soon know the warning signs and alert a responsible person.
And most of this technology is already being used somewhere, by somebody. Should you be investing in the IIoT too? To see what’s available, don your RFID-tagged hard hat and join our tour of Factory 4.0.


Smart energy management systems track fluctuating prices for electricity and gas, purchasing when they’re low and predicting when demand will peak. The system adjusts production schedules to take advantage of renewable energy. You can run energy-hungry processes on sunny or windy days, or postpone them when electricity prices rise, due to high demand and restricted supply. If your own solar panels and wind turbines give you more – or less – than you need to use, a Smart Grid will trade with neighbouring businesses on your behalf, keeping costs and carbon emissions low.


Once known as rapid prototyping, using a digital design to build up a solid object in layers is now a mainstream manufacturing technique. Creating custom components on demand is easy.

If you can design it in a virtual space, there’s a machine that can create it out of powdered polymer, metal, or even living cells. Combine different materials in the same process to create an electronic circuit within a solid object. Or, like NASA, create a bimetallic rocket igniter. You could even scan an existing object and recreate it with a 3D printer, with obvious implications for Intellectual Property.


The IIoT turns every point in your process, from batches of raw material to truckloads of finished goods, into a data point. As each data point updates itself in real time, your virtual business shadows the real thing. Cheap and discreet electronic RFID tags in every product let the automatic stock control system track each item from production to sale. Stock is less likely to go astray en route from Shanghai to Shenfield. And any irregularities in manufacture are traceable back to the source of the problem, so you can solve the issue in minutes, not days.


Yes, machines can increasingly work without human supervision. In fact, sometimes it’s the machines keeping a watchful eye on their human colleagues. Not only do smart ID tags mean no more manual clocking-in, analytical software could detect unusual patterns of behaviour that may be a sign something is wrong. Thanks to RFID tags in safety helmets, cranes can detect the presence of a fragile human head in the danger zone, too close to a swinging hook. Motion capture and activity sensors spot when somebody is tiring or losing concentration. And soon, facial recognition technology will be able to detect when an employee is unhappy – or maybe just annoyed about this constant surveillance!


Simulators are already used to train pilots, surgeons and drivers. Thanks to Virtual Reality (VR) and Augmented Reality (AR), onsite training now uses simulation too. Staff members can practise in a virtual environment, or augment the real environment with information, instructions and controls. Haptic interfaces even simulate the sensation of touch, letting you feel the resistance of a rusty bolt or the squelch of Swarfega. Trainers many miles away could interact with trainees on location, seeing what they’re seeing as well as what they’re doing. Cheaper than sending them on residential courses, and safer than letting them loose on the real thing right away.


Sure, big manufacturing plants have had industrial robots for years. Now smaller factories making lowvolume, high-precision goods are opting for co-bots. The new generation of mechanical workmates are designed to learn new tasks from human colleagues and then work safely alongside them. “It’s an ideal fit for the lower-volume, higher-mix environments which define the vast majority of manufacturing tasks today,” say Rethink Robotics. Their Baxter robot can reproduce actions that a human workmate has shown it, by demonstration or by manipulating the robot’s arm through the required motions. Baxter learns by imitation, no programming required. But do they ever take their turn on the tea run?


Spotting tiny abnormalities or defects is slow, repetitive and dull for a human being. That makes it the perfect task for an AI programme based on Machine Learning (ML). Fujitsu, for example, works with Siemens to analyse scans of 75m-wide wind-turbine blades faster and more thoroughly than the unaided human eye. Researchers have even tested ML algorithms with an electronic nose to speed up quality control of olive oils. Downtime Analysis also uses ML to analyse data when machinery stops working. By classifying downtime periods by cause – equipment failure, starved of materials or blocked from sending completed work onwards – it helps predict and prevent future incidents.

Take-out: industry 4.0 Is not a kit of parts, it’s the confluence of many different technologies. What will really make the difference for your business?