In a manufacturing facility continuous operation of the production line is vital. The entire company rides solely on the output of those lines. Even a company the size of General Motors fails if it stops producing cars.
When Microsoft Windows began gaining acceptance, the industrial market (with technology that has always lagged the commercial market by 3 years or so) began to allow Windows based applications into its plants to monitor and control production. The reliance on these systems has continued to grow to the point that now it is almost impossible to visit any factory that does not have some type of Human/Machine Interface (HMI) software throughout. PCs on a factory floor run 24 hours a day, every day. If the PC is essential to successful operation of the assembly line, then a down PC and its HMI can commonly cost a company anywhere from tens of thousands to millions of dollars every hour.
These large HMI packages carry around many man-centuries of code. New releases, especially major reworks, are viewed with healthy skepticism, as failure in HMI software simply cannot be tolerated. If your word processor has some problems, your letter may go out late or in the wrong font. If your industrial software has a train wreck, your plant shuts down.
Every manufacturer has the same problem - maintenance of PCs scattered throughout the factory floor that are required to run all of these interfaces. Plant operators are simply not qualified to keep PCs running, and the IT department avoids a trip to the plant floor if they can possibly help it. A much better solution is to move the actual computing power into the IT department and, following the mainframe paradigm, put just the operator interface on the factory floor. But doing this while maintaining the Windows interface requires some major rewrites. As we have seen, modifying legacy code to support this type of operation, either with traditional client/server or with a web interface, is almost like starting over. And even if software companies put the effort into the upgrade, the customer is often left with older graphics that cannot be converted easily, and is forced to run what he considers an unproven version of the application.
It is not uncommon today to walk into a fairly modern plant and find PCs on the factory floor running Windows 3.11. Not because it is better, but because they couldn't take the time to update the system, and passed what could be called the "update window". If you don't keep a system constantly current, the technology that you are using is no longer supported, the window closes, and you are stuck with a boat anchor.
But the real disaster is looming. Soon their 3.11 PC will fail. And because the window has closed, it cannot be replaced with an identical system. Even if they have another PC ready to go they are simply postponing the inevitable. There will be a time when that system must be updated. Not only will this mean new development expense (we are, after all, outside of the update window) but it will mean lost production.
Thin Clients give factories something they can migrate to right now that will solve their PC reliability and upgrade issues. On a true Thin Client, complex and specialized industrial software works just like it does on currently running PCs. In fact, the switch from distributed PCs throughout a plant to Thin Clients is possible without the operator even knowing the difference. One day his display is driven by a stand-alone computer, the next day by a Thin Client. But the plant manager notices. If a broken water pipe shorts out the Thin Client, all he has to do is take a new one and plug it in. Not only is all of the software ready to go, but the display is exactly where it was before the accident, and the application never even stopped running on the server.
A major benefit that is seen by the migration to Thin Clients is the reduction of long-term maintenance costs, often called Total Cost of Ownership. Software and upgrades are only installed on the server instead of on every PC in the plant. Maintenance of the Thin Clients is almost nonexistent, because they have become as easy to replace as a sensor. Backups and virus protection are only necessary on the server.
One more issue faced daily by factories is the cost of an industrial PC. Office PCs are now under $1,000, but an industrial PC will cost three or four times that much, and will have at best last year's technology. While it is difficult and expensive to make a PC that will operate in an explosive atmosphere at 110 degrees, it is very easy (translates to inexpensive) to make a 5V DC powered Thin Client with no disk drives for the same environment.
At ACP, we are excited about the future of industrial Thin Clients. We took what is a very good system (Thin Clients running on Microsoft Windows 2000 Server) and made it a great system, tailoring a Thin Client network to the specialized needs of the industrial and manufacturing workplace. ACP ThinManager provides for instantaneous, automatic failover to any available backup server if the client detects problems with the primary server. ACP Enabled Thin Clients can support industrial specific I/O (even very high speed serial data) and user devices directly from the client, and have an automatic configuration option for any new clients added to the network. Because manufacturers cannot afford to be tied to a single supplier, ACP Enabled Thin Clients come from a number of different manufacturers - there are about 40 different models to choose from. These features and more are the reason that we have generated such interest in the industrial market, and this interest is why we believe the future for Thin Clients, especially in manufacturing, is very bright indeed.
For more information on ACP Industrial Thin Client computers, please visit our web site at http://www.thinmanager.com
To sign up for the E-mail newsletter go here: ACP newsletter signup
For an archive of past newsletter articles go to: ACP Newsletter Archive