Not all that long ago, the family of industrial machine builders busily and, in general, happily designed proprietary control and communications systems for their machines.
These application-specific systems didn't carry a lot of overhead for unnecessary functionality, and the communication between devices, HMI, and controllers was heavily based on home-grown code. The builder had control over his costs, his architecture, and a relatively captive audience of customers that had to buy the entire system.
Customers weren't all that unhappy with the proposition. If something went wrong, they knew exactly who was responsible. Sure, the customer could have an integrator build a control system instead, but the problems were similar, and now there were two factions to deal with if problems arose.
The fabric of this universe began to unravel as open standards architecture began to bully its way from the office onto the factory floor. Machines now had to spit out operating data for further review and manipulation by the enterprise domain. Commercial Ethernet and Internet capability and Windows-based application hooks were showing up in RFPs.
Machine builders saw little upside in this for them. They had to learn a new architecture, component selection was now more complicated, and they had to send suddenly inexperienced technicians all over the planet to support this stuff when things went haywire.
Cut to the chase. To be precise, cut to the part where the light bulb went off in the collective heads of many machine builders, when they realized that maybe, with smart sensors and TCP/IP, there was something in this for them, too. Using open standards technology meant they could predict, troubleshoot, and diagnose problems across the country from a browser in their own office.
As this concept/benefit evolves, I'm always curious how it's actually being used. There's also the nagging question of who pays for it.
A panel discussion at Automation Fair late last year put some machine builders and users together to talk about on-machine issues, so I asked the panel to engage this subject.
Michael Hanssmann, vice president of the Controls Div. of semiconductor tool builder Brooks-PRI Automation, says that, within his industry, tools worth as much as $20 million each are being e-diagnosed from remote locations by specialists. "The machines are designed from the ground up to allow this to take place." says Hanssmann.
But, OEM and customer clearly differ about cost and benefit. "There's the argument that the equipment manufacturers benefit from reduced fly-and-fix," Hanssmann says. "So, the semiconductor manufacturers say it's the equipment manufacturers that should bear the cost."
Reverse the roles and you find, "Machine builders are saying [the end users] get a lot better up time and better throughput, so the users should be bearing the cost," argues Hanssmann. "I'm not sure where it's going to end up."
Gary Cash, director, controls engineering for material handling system builder FKI Logistex, reminds us that viewing this service as a real value-add is very much industry-dependent.
"Most of our customers have some interest, but they're not willing to spend any money," says Cash. "Look at the industry and you'll understand why. If your semiconductor equipment goes down, you're stuck. When the beer handling equipment goes down, we're all in trouble! But if your conveyor stops, you can still get the boxes to the door. You do it manually." He says they usually just wait until it breaks, and they're not pushing him to keep track of it.
This issue is a sort of maintenance-insurance argument. A lot of money goes into marketing and distribution. Maintenance is something no one really wants to spend money on. But, realistically, it has to viewed in the entire context of operational excellence
Sorting out the parameters that demonstrate the value is, and will continue to be, a work in progress that we'll keep track of for you.
E-mail Joe at firstname.lastname@example.org.