By Mark Devlin, IEN Staff
May 1, 2008 -- The editors of IEN compiled a fairly extensive list of predictions for our 50th anniversary back in 1983. That list categorically organized predictions by Factory of the Future, Robots, Sensors, PLCs, Computers, and Fiber Optics. All hot areas, indeed. In this, the 75th year of IEN’s service to readers, let’s dust off the crystal ball and look back at several of those predictions and how they actually turned out.
Unmanned and Automatic
The “lights-out” factory, in other words, was a 1980s vision of Roger Smith, then chairman of General Motors. He imagined robots and factory automation components so reliable that they could work in the dark without human intervention. Smith’s dream hasn’t yet come to fruition at GM, though it came true at Fanuc, one of GM’s suppliers of robots. According to a Business 2.0 article on CNN.com (See References), “...in one of Fanuc's 40,000-square-foot factories near Mt. Fuji, robots are building other robots at a rate of about 50 per 24-hour shift and can run unsupervised for as long as 30 days at a time. When they stop, it's because there's no room to store the goods. Trucks haul off the new robots, the lights are cut, and the process begins anew. 'Not only is it lights-out,' says Fanuc vice president Gary Zywiol, 'we turn off the air conditioning and heat too.'" While still a rarity even today, some plants crank out parts for days or weeks at a time, stopping only for humans to replenish supplies.
The Ubiquitous PC
Keeping in mind that the IBM Personal Computer was only 2 years old in ’83, a lot of pundits completely missed that the PC would become a pervasive world standard for computing, earning a place in the vast majority of homes, virtually every business on the planet, and both the office and plant floors of manufacturing companies. That said, many futurists (such as an author in Mechanix Illustrated) did, in fact, predict that there would be a computer in most homes.
When introduced, the IBM PC had a 4.77 MHz Intel 8088 processor and 5.25 in. floppy drive for about $3,000. Not exactly the bargain of the century, and easy to miss in terms of anyone realizing at the time that, 25 years later, something so underpowered and expensive could become a global standard of computing. At IEN, we missed that this platform would become the basis for many small- to large-scale manufacturing operations, in many cases supplanting PLCs and, in others, working with them. Some also believed then that the Mac would be the standard for testing and measurement, for example and, for a short time and on a limited basis, it was. It didn’t take long, however, for the PC's horsepower to increase, price to decrease, and platform openness to catch on—in the process, making it one of the most life-changing technology devices of all time.
There’s at least one general, computing-related caveat to the above. IEN editors predicted in ’83 that lasers would replace electrical circuits within computers. Just last week (on March 24, 2008), Sun Microsystems announced a $44M Pentagon contract to explore the viability of replacing constrictive electrical circuits with lasers enabling incredibly rich and fast communication between arrays of computer chips, a technology now called silicon photonics. Of course, this will be only for best-of-class supercomputers, at least until IEN’s 100th.
Fiber Optics
We made several predictions in ’83 relating to fiber optics and largely, those predictions were accurate. So accurate, in fact, that something once seen as gee-whiz science has become so common that fiber optic components can be ordered online from Grainger, for example.
The mindset back then, though, was such that everything needed to be physically connected via circuits, wires, and cables—either of the electrical or optical variety. No one foresaw wireless at a time when pretty much the only available wireless device (save for radio or television) was a 2-lb cellular phone with a half-hour of talk time and a $3,995 price tag. Today, thanks to the IEEE, anyone can setup a wireless home computer network for less than a few hundred dollars. Areas of cities, airports, hotels, and soon, even cars (thanks to Chrysler) have wireless, 802.11-based networking and Internet access. More importantly in the context of industrial equipment, we’re seeing the beginnings of wirelessly networked manufacturing—all made possible through the wonder of the Ethernet standard gone wired to wireless, and everything in-between (e.g., wired-in, wireless-out).
A Single Bus
The fieldbus wars raged in the 1980s. Arguably, more than 20 fieldbus "standards" existed back then, with each manufacturer pushing its own flavor of factory-device communication protocol—and their associated products, oftentimes locking users into devices from one manufacturer. In this context, IEN editors then predicted “The most important developments are likely to come in the field of compatibility.” We were half-right. The field’s been narrowed down. While the fieldbus wars are over, skirmishes continue taking place even in the land of Ethernet—where most suppliers are headed. While Ethernet seems a ubiquitous standard, things on the factory floor get a bit more complicated in terms of how and when signals are transmitted. Thus, suppliers have again developed their own flavors of tweaked Ethernet. It will take at least a few more years to see if any single, new standard wins—such as Ethernet/IP, EtherCAT, or PROFINET. We’re much farther along in terms of different devices communicating with those of different suppliers, so perhaps the market will decide that just a few options will work for their needs.
Robotics and Sensors
Many sensors used to be large enough to fit in your fist. Some still are, but that’s rare. Since the 1980s, sensors have predictably become more accurate, reliable, and repeatable, while also being considerably smaller and much smarter than their predecessors. Sensor-on-a-chip designs now easily fit on a fingertip, and the economies of scale have reduced costs such that sensors can now be embedded right into equipment before it ships. Even notebook computers now have two, three, or more sensors just to detect temperature, tied into the notebook’s cooling system to make incremental adjustments based on, for example, processor, hard drive, and internal case temperatures.
While sensors don’t require robots, today robots increasingly use sensors in end effectors, for example, so that gripping strength can be widely and precisely varied.
We predicted, for example, that sensors will be increasingly used to “duplicate human sensory faculties.” This has happened in spades, and the need for more human-like sensing has expanded far beyond touch to, for example, electronic noses that can sniff for bad or "off" scents 24/7/365.
Social, Ultraprecise Machines
We predicted then that factory floor equipment would become "sociable." Most pieces have, in fact, become networked with everything from sensors and controls to plant/corporate computers. Connected, however, doesn’t mean that they’re social. That’s coming, however; it’s just going to take a little more time for the typical factory to embody machines that not only converse, but make autonomous decisions based on the information in those conversations.
Also, we predicted that machining would evolve to the point where no final-inspection is necessary. In today’s reality, we’ve moved closer to that goal. Just as much QC work is required in some applications, less inspection is required in others, and in still others—in medical device manufacturing, for instance—hand-work is still required to finish the part after it’s been machined. We were on the right track, though—just a bit ahead of our time.
See the related Predictions articles for the future that’s starting now—and expected farther down the road.
