Remember the factory of the Future everyone dreamed about in the 80s? In the U.S. it's still more of a dream than a reality. Ironically, some developing countries may be closer to seeing the realization of that dream than those who originally dreamt it.
In this dream a point-of-sale system is connected to a warehouse which is connected to an on demand manufacturing plant operating in a lean—not a batch—mode. So if an item gets purchased at a particular location you have the logistical and manufacturing infrastructure to produce that one item and its replacement.
I spoke with Bill Torrens, director of sales and marketing for RMT Robotics, the other day, and he thinks the reason developing countries have a good crack at succeeding with automation where many in the U.S. have failed is that they have no entrenched systems standing in the way. There are some U.S. manufacturers trying to establish a global presence with logistics information systems but their warehouses and distribution centers are still mostly manual, although they are relying on islands of automation in their plants. Torrens told me he's seen third-world competitors to those manufacturers who are positioning themselves for a game of leap-frog on the plant side.
“Go into these manufacturers in the third world and it's like stepping into Oz,” he said. “You see German, American and Japanese manufacturing technology making parts and a zillion people pushing these parts around.”
U.S. manufacturing has been automated much longer than its third-world competitors but its warehousing is still just as manual, by and large. What needs to happen is better coordination of data and material flows between manufacturing and distribution. Torrens gives kudos to Kiva for its success in installing robotic goods-to-person systems in the DCs of many high-profile U.S. companies—most recently Amazon, which liked this flexible technology so much it bought Kiva. In a sense it was easier for Kiva to conquer the world of U.S. warehousing with its concept because that world is still a blank slate where automation is concerned. U.S. manufacturing is a different story, he says—one which his company sees as the next opportunity for robots.
“The reason KIVA is successful is that in most warehouse applications, there's nothing,” Torrens said. “There may be a little bit of conveyor, but warehouse applications in 2012 are mostly manual. It's the last frontier of automated material handling. The manufacturing environment is different. It's not a clean slate and you can't just rearrange everything. You're getting involved in incumbent layouts which are difficult to adapt to. You just can't wipe the slate clean.”
U.S. Manufacturers are still trying to achieve a return on investment from their plant automation, and they expect to do that with savings. For Torrens, that's the wrong approach. Technology should be seen as a money-making proposition, not a money-saving one. Take automated guided vehicles (AGVs), for example.
“Generally speaking labor displacement with AGVs often represents only 60% of their return on investment,” he continued. “The remainder has to come from efficiency improvements. Why do a lot of manufacturers not apply AGV technologies? Typically they might have tried AGVs for fixed path and old style traffic management and that technology didn't work for them.”
That's the opportunity Torrens sees for his company, which makes Autonomous Mobile Robots (AMRs). These machines have four characteristics: first is feature based localization, meaning each one knows where it is in relation to other objects; second is dynamic path timing, meaning the robot decides on how to get where it's going autonomously and navigate around obstacles; the third is autonomous traffic management, where the vehicles “talk” to each other and orchestrate their movement through the building so that they most effectively get where they need to go without traffic jams; fourth is power management, meaning they know when they need to re-charge.
So how does something like this make money? Not just by replacing labor, but by making existing automation 15-20% more efficient.
“The traditional â€˜transportation' methodology of AGVs is limiting and prevents the automation it's servicing from reaching its potential,” Torrens said. “It wants five items and you deliver 50. It doesn't want 50. It wants five items ten times a day. By delivering a pallet load of 50 you're actually hindering the efficiency of the operation. That's the opposite of lean. If you can make the automation you've invested so much in more efficient, and it produces 10% more because of it, you make money. That's where the mindset of AGVs needs to turn.”
I know what you may be thinking at this point. Torrens is a sales guy, and that's exactly what he's doing. Selling. But what he's talking about is a piece of a larger phenomenon taking hold around the world: machine-to-machine (M2M) communication. A new report, Rise of the machines: Moving from hype to reality in the burgeoning market for machine-to-machine communication, sponsored by SAP, examines the business models behind successful M2M applications across sectors. In it, some of what Torrens described is already happening with a variety of different technologies.
Logistics firms such as UPS use M2M in their over-the-road vehicle fleets not only to optimize driving routes, but also to provide live package tracking information for customers. Within the four walls, a New Zealand-based dairy company has set up autonomous forklifts in its warehouse that can work around the clock, with far fewer accidents and reduced wear and tear.
The report also cites examples outside logistics. U.S.-based Progressive Insurance sets rates based on actual driving habits. Those habits are monitored by technology like GM's OnStar system which provides services ranging from automatic collision notification to remote door unlocking. Then you have TomTom, a satellite navigation provider, which automatically tallies traffic information from millions of users to set better routes for other drivers.
There are still plenty of roadblocks to M2M. Although costs are coming down rapidly, operators and systems integrators still have to standardize technology platforms and develop open protocols to allow for tighter integration between sensors, devices and other hardware. And user concerns about data security and access still need to be addressed.
Big Brother is alive and well. Just keep your machines from talking to him.