Penny wise and pound foolish
A few appropriate costs early in a project can create major savings down the road and avoid a whole lot of aggravation in the process.
I’ve just finished helping support a project startup that has gone rather well in spite of some decisions to save cost at the start. I continue to be amazed time and time again that companies underestimate the impact of cutting the wrong costs, simply because they don’t know what the project really should cost before they go for funding. The project in this particular case is actually the follow-on to the first phase of a system migration from one DCS platform to another. The customer’s old platform had been in service for a very long time and was becoming unsupportable when things failed. Since any migration was going to be a rip-and-replace, they decided to investigate what was now available in the market place and chose not to stay with the original installed platform vendor for technical reasons. Phase one was done by the vendor of the new DCS, but to save cost the configuration was off-shored. This might have worked if the customer had developed a really good scope, but for reasons I do not know, they did not.
This resulted in a lot of issues, including dissatisfaction with some of the code they received, overloaded controllers, and operator workstations that needed frequent rebooting. The next two phases have cleaned up some of the legacy code issues, but progress has been hampered by the system hardware issues and lack of a really complete scope. They planned to have a complete tight scope but the resources weren’t available internally to finish it in time and they chose not to go outside to get it done. There must be a variety of reasons but I’m sure money was on the list. They could have eliminated the problems we are now dealing with if they had engaged a team experienced with front end loading (FEL) processes at the outset.
In today’s business environment companies are leaner than they have been at almost any time in the last 40 years. That doesn’t leave much bandwidth for staff to take on tasks involved in the migration of an entire facility from a 30+ year old system to the latest technology. Just ferreting out where all the field devices are, what their condition is, and whether or not they are still being used is no small undertaking. Cutting corners there will certainly cost in the long run.
On another recent migration project with a different customer we discovered during the cutover that we were de-terminating and re-terminating wires that didn’t even have instruments connected. We started doing it because the loops were still in the DCS database and wired to the I/O. So much for keeping up-to-date. There were even a few that had been migrated to a fieldbus device and were being monitored by the new system but the old analog wires were still connected to the old system, so they got moved. Field labor is a lot more expensive than office labor so everything that can be done to eliminate it will produce savings, provided it’s done effectively. The people doing your FEL work need to have a robust methodology that is used consistently to produce verifiable results.
Notice I didn’t say anything about aiming for low cost results. Let’s face it, the low cost approach may be perfectly fine if the results are there. It’s like the project outcome triangle: You can have it fast, you can have it good, you can have it cheap–pick any two because you can’t have all three. For this project, taking extra time up front for a really good estimate would have identified that the controllers were most probably going to end up more heavily loaded than is good. Given that the cost of a controller is relatively small in the whole scope of things, buying a couple more would have paid dividends in the long run. It would have also pointed out that not spending a few hundred dollars more for the workstation PCs to get more memory was not really a cost saving if they lock up and the operators are rebooting them several times a day. But the worst cost saving measure was buying cheap chairs for the operators to sit in all day. They probably won’t last six months. At least I know firsthand that it caused the most uproar. Never skimp on operators’ chairs.
This post was written by Bruce Brandt. Bruce is a Technology Leader at MAVERICK Technologies, a leading automation solutions provider offering industrial automation, strategic manufacturing, and enterprise integration services for the process industries. MAVERICK delivers expertise and consulting in a wide variety of areas including industrial automation controls, distributed control systems, manufacturing execution systems, operational strategy, business process optimization and more.