Monday, 6 June 2011

Remote Troubleshooting

There was a difficulty last week with ongoing attempts to remotely troubleshoot problems with an effluent treatment plant in India. I always like to troubleshoot things in person if at all possible, and the difficulties we are having remind me of why.

The essence of this particular problem is that the client organisation believes that certain interventions have clearly improved matters, when the little data they have collected suggests otherwise.

It's always tricky to tell a paying client that they are mistaken, especially when you are telling them that the truth is less favourable than their version of events. It is however in my opinion a professional duty not to allow them to live in a fool's paradise.

It is always better when troubleshooting to take nothing for granted, to establish the validity of every important parameter personally. Experience suggests that somewhere in the pre-existing set of assumptions is the duff one which has prevented resolution of the problem prior to your arrival.

Teasing out the bad premise is as often as not the troubleshooter's most difficult task. This is often not difficult technically, but it can be politically troublesome. No-one noticed before you arrived something which can seem obvious with hindsight, and as no-one wants to look foolish, getting the data you need can be very tricky.