Why (most) industrial IoT platforms suck – and solve nothing by themselves
Sometimes content gets left on the cutting room floor, and sometimes it does not even make the edit. And often that content is excellent, and deserves a feature all of its own. This is a (delayed) retelling of a briefing Enterprise IoT Insights had with an ‘OT operator’ a couple of months back in Sweden, at the European leg of PI World, a bi-annual gathering of industrialists working with PI System, the IT/OT data integration suite from California firm OSIsoft.
The event is notable on the conference circuit because it does away with the ‘future selling’ that characterizes most other tech summits. Instead, it was all business; a private affair detailing the minutiae of industrial IoT mechanics, mostly told by the enterprises themselves. In Gothenburg, in September, the OT set sought to put the IT industry back in its ‘magic box’, as we have recorded.
Part of the press briefing – picked up and put back into the edit, here – considers the mess of over-the-top data integration platforms, bundled for marketing as ‘IoT platforms’, offered by technology firms. There are 400 so-called IoT platforms in the market, we hear; an analyst, also in the briefing, reckons on 700; Enterprise IoT Insights has been quoted no fewer than 1,200.
Either way, there are a bunch of them out there, and most of them suck (inevitably). And all of them, we hear, are useless on their own. Why? Because they are made by software companies, rooted in the IT sector, and presented as a panacea for industrial IoT – a way to unite shop-floor and back-office data in a single system, and press ‘go’ on industrial transformation.
“Platforms are solutions without problems.” We are in conversation, in a briefing room in Gothenburg, with Conversión Magro, deputy chief executive at Alizent, a subsidiary of French firm Air Liquide, the world’s largest supplier (by revenue) of industrial gases. Alizent was founded in 2017, uniting a trio of digital services companies owned by Air Liquide, as a vehicle to marry its parent’s industrial knowledge with software and consultancy services.
Rather like OSIsoft, which supplies underlying databases for industrial sensors and systems, it claims rarefied crossover appeal, between the IT and OT crowds. Unlike vendors tackling the IoT challenge from the other side, it gets industrial culture, it says, in a way that lets its go to the heart of the matter, and pull at the real levers of industrial change.
“We are not about technology; we are about putting technology to work for the industry. That’s the difference,” says Magro. “We don’t talk about projects. We don’t talk about platforms. We talk about solutions. You have a problem, and we find the elements [to solve it] – and maybe some of them are not so glamorous, or so ‘IoT’. But they work, and they work globally.”
The difference, he explains, is technologists have all their eggs in one tech basket; they know no other way, he says. “All of the value is in the platform, in the technology. They might put a finger over the line, to collaborate with domain experts. But they won’t go further than that; they will not deliver those [industrial] solutions. We are the opposite; we are entirely in the industrial space – our whole body is over the line.”
Magro started as a software engineer in 1998, he confesses. But his journey with Air Liquide, moving into senior management positions before helping to create Alizent in 2017, has taken him beyond his roots, and entirely across the IT/OT divide. IoT platforms, he reasons, are the products of IT minds addressing OT problems from the outside. “It’s how I would have solved a problem 12 years ago, when I was still a software architect,” he says.
“It’s what my engineering mind would have told me to do: separate out the part of the problem that I do not control – which is so huge; which covers the industrial part, the process part, the organizational piece, all the complexity of using digital in industry – and address the part I do understand, as an engineer. Which is the connectivity, the protocols, the forest of different brands, and all the devices I have to connect – and to try and generalize that.
“But you cannot think that by solving that part of problem, which is the technical wiring and plumbing, you are solving the rest of it as well. You are still really far away, still, from solving the other [bigger challenge of industrial IoT]. Can this piece contribute to the wider solution? Of course it can. But you need imagination, as well. You need to bridge the gap.”
And bridging the gap takes an unknowable amount of time – “a trillion man-days,” he says. Crucially, it also takes tremendous vision and desire on the part of senior management to make wide-ranging and effective digital change.
A decent IoT platform, on its own, only provides a window into industrial systems and processes; questions around what and how to change operations go much deeper, and will not be answered by IT technicians. Coders write code, sure; but they need hand-holding when it comes to first inputs and final outputs, and how these flow across an organization seeking total reinvention.
Magro puts it best. “If platforms get to a mature point, then they will shorten that gap, to do something useful. Because normally you have to customize it. But that is not the case today. Because the market is still being defined. And if it costs you a million man-days, at some point you wonder [what the point is]?
This is an excerpt from a new editorial report on co-creation, co-innovation and co-configuration, out next week.