Manufacturing is about to cross the physical-digital divide
If you’re curious about what the future of manufacturing holds for years to come, you won’t necessarily find it by touring an assembly line or a cutting-edge chip fab. Nor would you appreciate its potential by sitting down in a new car or powering up an Intel-powered laptop. It’s not really so much about the thing, as impressive as those may be. The future of manufacturing is more about the convergence of the physical and digital worlds into something entirely new—and with huge potential economic value.
Physical manufacturing is going digital—and the sum is greater than the parts
As physical products today take on exponentially more complexity in the age of IoT, 3D printing and mass customization, technology has emerged that can unlock vast economic potential by virtualizing every phase of the manufactured-product lifecycle—from initial design concepts and prototypes to functioning “smart” products to complex installed systems out in the field. As physical products become more enmeshed with software, sensors and their intrinsic data streams, their capabilities are expanding at a nonlinear pace.
Technology has evolved to a point where predictive models can foresee precisely when, say, a welding robot will fail, under what conditions and after exactly how many hours of operation. They can predict what specific shape of turbine blade will last the longest during a rough winter at an offshore wind farm. They can tell rocket builders what type of lightweight materials will launch safely without the need for physical tests—saving billions of dollars, potentially, and perhaps human lives.
While over the last couple of decades tech innovation has revolved around the explosion of all-digital products, companies and services, a new wave of invention is coming to what many see as a re-imagined manufacturing economy, one in which these digital capabilities are built into every aspect of the manmade world, from the front end (simulated design and prototyping), the middle (physical manufacturing) and the back (operational monitoring and optimization). As Harvard economist Michael Porter explained recently, “smart connected products can ultimately function with complete autonomy. Human operators merely monitor performance or watch over the fleet or the system, rather than over individual units.”
The new world of simulation
While simulation software has been around since the days of the Apollo space program, its convergence with IoT capabilities has greatly expanded the possibilities—to the point where it can create ideal, buildable designs beyond the reach of human skill. The Silicon Valley startup Nebia, for instance, tapped into new simulation tools and models to design a showerhead that its engineers tried for several years to perfect. Simulation and data modeling showed Nebia how to precisely design and manufacture thousands of micro-nozzles in such a way to slash water consumption by 70 percent while maximizing heat retention. The real magic, though, is in the bottom line—Nebia’s development time improved dramatically, as did production costs.
What once took years for engineers in the field to test can now be accomplished in hours or days on a screen with reams of data available and unlimited computing power on demand.
The additive revolution
Mass-scale commodity manufacturing in places like China and India is on the decline—and not just because of economic trends. IoT-driven manufacturing is ushering in a new era of customization that leans into an on-demand model, where you can not only simulate a digital prototype but customize every aspect of the build—and every IoT-enabled part—without sacrificing scale.
Imagine you’re a niche maker of surfboards and you’ve come up with a cool new design, but you want to make sure the board has certain physical properties so it doesn’t break, and it’s built with materials you can afford. Until recently, you were stuck with a legacy model of getting to market—after building a few prototypes, you might ship it off to China for manufacturing. By the time you work out all the kinks in the process, it could take several years to get to market.
With the rise of 3D printing and the IoT, a new approach called additive manufacturing is flipping the process. Designers can create the optimal design; simulation modelling will tell exactly how that design will look and behave, and what materials to use in the manufacturing process. Then, instead of going overseas, manufacturers can 3D-print their own prototypes or find a local partner who can do it for them—just upload your digital files and your partner might turn around 1,000 skateboards in a couple of days. Turnaround time becomes near instantaneous, costs are minimized and—an added bonus—you’ve kept your IP out of the hands of a foreign factory.
Predictive modelling and monitoring
Manufacturing giants such as Schaeffler create products used in asset-intensive industries worldwide—automotive, transportation, energy and industrial products. Yet all of these physical things are fast evolving into dynamic digital systems, enabled by myriad sensors that track every aspect of their operations. They throw off so much data, in fact, that engineers can now create “digital twins” of these systems—virtual doppelgangers of the real thing, operating virtually in its unique physical environment, over a range of conditions and time. That has huge implications not just for operational efficiency, but for market innovation.
With digital twins, engineers can analyze and optimize the performance of products in real-world operating conditions. They can see not just what is happening at any given time, but why. They can speed up simulations, pinpoint when breakdowns will occur (and why), and reduce the cost and risk of unplanned downtime. They can also predict the future performance of a component as it operates within a larger system—a wing on a plane as it travels from San Francisco to London; a rocket engine as it undergoes the violent progression from launch to stage separation; an office building as it manages power, energy and HVAC systems through the course of a day; and extending much further, a driverless car, navigating city streets at rush hour with hundreds of other virtual cars on the road, each one represented by complex sets of very real physical and operating data.
Looking to build a new wind farm? In the not-too distant future, all the critical groundwork to consider that decision will be virtual. Start with the wind turbines—the specific type and shape, the power generation capacity and the climate factors that affect seasonal operations. You’ll have at your fingertips all the sensor-driven data that logged the service records of people who have climbed the tower and done service checks. Aggregate all of this data within the “twin” and see if your system will work from every angle—how efficient it will be across all weather conditions; how much downtime you will need to account for; what tradeoffs to consider in terms of long-term operational efficiency.
The ultimate promise of physical manufacturing merging with the digital universe isn’t that it can produce an energy-efficient shower-head that human engineers couldn’t figure out using traditional development methods. It’s that it can peer a bit into the future and allow everyone from designers to CEOs to ask—and then act upon—the ultimate what-ifs: What if we redesign this product to exploit nascent market opportunities? How can we apply new manufacturing techniques or incorporate new materials to improve product quality? The IoT will only accelerate the potential and benefit of these changes in the manufacturing industry.
Learn more about Watson IoT and Industry 4.0.