Just less than 25 years ago, the first dot matrix indicators giving real time information appeared on the underground. And what a novelty they were. Now, we can't do without them or without the other pieces of real time information we get on the web, on twitter and elsewhere.
But they do lull us into what is sometimes a false sense of security because we believe they are accurate. At Chilham, where I wait for the train each morning, the indicator usually says my train will arrive 'on time'. Sometimes it does. Other times it can be anything up to 5 minutes late. So we all get out our smartphones and consult our various apps and websites to decide what 'on time' means today.
Even the web based information isn't perfect and a frequent outcome shows the train has left Canterbury West 2 minutes late - having been 2 minutes late up the line - but is still forecast to be on time at the next stop, Chartham, only 5 minutes run from Canterbury. How anyone seriously expects the train to shave 2 minutes off a 5 minute run escapes me.
On other occasions we see 2 minutes late at Canterbury West, on time at Chartham and 2 minutes late at Chilham. And the train still turns up anything between on time and 5 minutes late.
Well, you may say, what's 5 minutes between friends (or even passengers, sorry, customers). According to the definitions, even 5 minutes late is still, strictly, 'on time'. But it does matter because it damages the confidence in the information provided overall.
On occasions of more widespread disruption, the system collapses entirely and perhaps that's not surprising. But it still produces false positives. During the snow, last winter, we regulalrly saw trains posted as 'expected on time' and time would come and go with no sign of any train. And then that train would dissappear off the display. All the time, a more detailed interrogation of the website showed that the train was not running at all, had never started and was, in effect, cancelled. (I wonder how that shows up in train performance statistics.) But most passengers won't do that more detailed interrogation, and why should they need to?
Vince Lucas helpfully told me that the system at Chilham was put in on the cheap on the basis that something is better than nothing, but that it couldn't be as reliable as information at bigger stations. He is almost certainly right that something is better than nothing, but our expectations need to be managed better if we can't wholly rely on the information provided.
It's a hoary old question if bad information is worse than no information, but we have to give much greater priority to minimising errors and discrepancies and thinking how we get across the limitations of the systems. A pretence of accuracy which simply damages any confidence in the system isn't the right answer.
No comments:
Post a Comment