Your car has yet to reach its first birthday when the transmission seizes up unexpectedly. A mechanic informs you the culprit is a hairline crack on a gear deep inside the engine block. You haven’t been in any collisions, so how did a faulty part end up in your new vehicle?
To avoid irate customers and costly repairs of cars under warranty, auto makers like Nissan and parts manufacturers such as Germany’s ZF Group are turning to Acerta Analytics Solutions, a Kitchener, Ont., start-up that employs data analytics to identify components at risk of failure. Using sensors, the company’s system scans a transmission component, for example, for patterns of vibration or material irregularities and compares them to previous defect reports. A match indicates a part is potentially faulty, so it’s yanked off the line before it can be installed, explains founder Greta Cutulenco. And this all happens remotely: Acerta clients simply subscribe to the system, which analyzes a continual flow of data transmitted from factory floors.
The 20-employee company is but a tiny cog in a vast machine of connected technologies collectively dubbed Industry 4.0. Following transformations ushered in by just-in-time delivery, automation and supply chain integration, this next revolution in manufacturing will be driven by data analytics, artificial intelligence and wireless connectivity. The aim is to harness all the retrievable but mostly ignored data that factories generate – such as patterns in equipment breakdowns or past supplier records to project delivery lead times – and use it to improve productivity both in the plant and throughout the supply network. Researchers put global spending on these technologies at more than $200 billion (U.S.) in the next five years, with automotive, aerospace, and food- and beverage-processing companies expected to spearhead its adoption in Canada.
“Some companies are using less than 1% of the data they’re collecting,” says Cutulenco, a software engineer who formerly worked at Magna. Aware that modern vehicles pulse with electronics, she reckoned the output could be analyzed to home in on patterns. It was a struggle to get auto-manufacturing “elephants” to see the value, she says, so initially she and her partners drove around southern Ontario’s industrial corridor with hard drives, literally downloading files from vehicle on-board diagnostics to demonstrate their concept.
Industry 4.0 is all about squeezing value from this kind of machine-gathered data. Say a sensor embedded in a machine identifies a part that’s showing signs of stress. It can notify a cloud-based system that checks if the part is already on site, and if not, seeks quotes for replacements, places a purchase order, and times the delivery to coincide with a technician’s availability and an optimal time to take the equipment offline. When the part arrives, “it’s the first time a human will interact with that system,” says Sam Masri, chief operating officer of SAP Canada, an enterprise software developer working on such solutions.
Masri points out that the lower cost of sensors allows manufacturers to use technologies such as machine learning and the Internet of Things to aggregate performance data; an earlier generation of industrial robots were programmed to perform specific tasks but weren’t networked. “A few years ago, the high cost would have made it impossible to add a sensor to every single component in a factory,” he says. “Now it’s full speed ahead.”
That may be an overstatement. While pilot projects abound and, according to a recent McKinsey & Co. survey, 68% of industrial firms see a move to digital manufacturing as a priority, companies worry about keeping all that data out of hackers’ hands and having the right talent to run a smart factory. What’s more, Canadian manufacturers seem to be lagging behind other countries in technology adoption. Sales of industrial robots in Canada actually fell between 2015 and 2016, according to the 2017 World Robotics Survey. While South Korea has 631 robots per 10,000 employees, in the Americas, the figure is just 84.
But companies have some powerful incentives for making these investments. John Sicard, CEO of Kinaxis, a supply chain automation firm in Ottawa, points to the relentless squeeze on product life cycles, especially in sectors like consumer electronics. Facing shorter sales periods and smaller runs of more customized goods, manufacturers can tap the new technologies to make the most of their production schedules and rapidly retool assembly lines for the next generation of products. “You have a very narrow [sales] window,” he says, “and you have to make sure you capture every sale possible.”
The data-driven applications are converging with other emerging technologies, such as 3-D printing and autonomous cars. For years, guided vehicles – basically, high-tech skids that can be steered remotely – have been gliding around factories, hauling trolleys laden with parts to workstations. Now, some plants are deploying sensor-equipped autonomous vehicles that not only navigate on their own but can also independently respond to obstacles on the factory floor. One of the early leaders in this arena is Otto Motors, a division of Clearpath Robotics, part of a growing cluster of Industry 4.0 firms in the Toronto–Waterloo corridor.
Beyond boosting productivity, smart factory tech may bring radical changes to manufacturing business models. UPS, for example, is experimenting with 3-D printing so it can not only deliver but also produce replacement auto parts for customers. Tesla, meanwhile, downloads software fixes to its vehicles to deal with some repairs, a system that cuts repair shops and dealerships out of the equation.
As was the case with earlier productivity-enhancing innovations, the gains will have an impact on the manufacturing workforce – a point not lost on policy makers. According to a 2017 federal advisory council report, almost a quarter of all factory tasks may be automated by 2030, a development that would have huge implications for the 1.7 million Canadians who work in the manufacturing sector.
Jayson Myers, who heads Next Generation Manufacturing Canada, a group pushing to make Canada a global leader in advanced manufacturing, says factories of the future will require employees with skills in data science and programming who can work collaboratively with the smart robots and other technologies deployed on the shop floor. Indeed, high-skill (and potentially high-paying) factory jobs may well offset positions lost due to growing automation. Amazon, for example, continues to hire huge numbers of new employees, even as it replaces more and more warehouse and clerical positions with robots.
Could factories someday become fully automated? In Shanghai, a massive new container port has already eliminated 70% of the labour costs associated with older facilities. Ross McKenzie, managing director of WatCAR, an automotive industry think tank at the University of Waterloo, conjures up the prospect of so-called “lights out” factories with no workers at all. An autonomous truck simply backs up to the loading bay, and robots load or unload the vehicle. “Inside the factory, there are no humans,” he says. “That’s the ultimate factory of the future.”
Editor’s note: An earlier version of this story misspelled Sam Masri's name as Misri. This online version has been corrected.
More from Factory of the Future
A better way to predict meat safety problems
Your underwear is watching you
Intelligent robots customize car seats on the fly
A digital approach to small-batch beer