Solid is about the intersection of real and virtual — the idea that, through sensors, networks, and intelligent machines, information can move fluidly between software and the physical world. It’s easy to see the technical implications of that intersection — thermostats that adjust themselves and cars that can drive autonomously — but there’s also a crucial management implication as well. Just as design can be automated and optimized if it’s encapsulated in software, a company’s operations can be made much more efficient if they’re modeled digitally before being executed.
Jamie Miller, senior vice president and chief information officer at General Electric, calls that “IT meets OT” and sees the combination changing her industry. “When you take these two disciplines that used to be separate and combine them, you can start to approach engineering and design differently, operate workflow differently, make factories brilliant.” It’s a philosophy that GE uses internally and builds into the products it sells.
Companies like GE have a lot of data — digital designs for manufactured parts, human-resources records, work orders from customers, service manuals — and this data tends to converge on human operators. A field technician might receive a work order to fix a wind turbine, visit the machine, consult documentation, call a colleague for specialized advice, order a replacement part, and finally make the repair.
That kind of operation becomes vastly more efficient when it’s dispatched by software. Sensor data, transmitted over a network and combined with predictive software, might obviate the first few steps by proactively forecasting a mechanical failure ahead of time. Then, a combination of data from other sources could speed the next few steps: supply-chain software finds the needed replacement part and sends it to a technician, selected for her specialized expertise and availability, who is dispatched to the job site with specific documentation. The entire process can be optimized at a high level by software that sits above these many individual systems.
That’s a management mindset that draws on both the broad IT databases that companies have been building for decades as well as the controls and processes that touch low-level operations — GE’s manufacturing models go all the way down to individual toolpaths in factories.
Miller describes her “brilliant factory” model in four pillars. Virtual design and manufacturing involves a series of all-encompassing models that link design to manufacturing. They’re able to immediately predict how any design decision might affect manufacturing cost and supply-chain throughput. It helps that every machine is instrumented with sensors — the second pillar — because the impact of change can be measured and built into models. Applying these models at wider scope brings about the third and fourth pillars: optimizing the factory itself and optimizing the entire supply chain. Together, these pillars create an approach that could yield important outcomes, such as reduced downtime and significant cost savings.
This post is part of a collaboration between O’Reilly and GE exploring the convergence of hardware and software. See our statement of editorial independence.