I remember one boss who, once in a blue moon.

I remember one boss who, once in a blue moon, would walk around the office; whenever he crept up behind me I was invariably writing a shopping list or was on the phone to my mum.

Such surveillance did not improve my behaviour, though it did increase my sense of injustice. To have been monitored all the time – which would have put the shopping list in the context of otherwise diligent behaviour – would have been a vast improvement.

In most offices a raft of mainly pointless, cumbersome tools are used to assess performance, including “competency matrices”, appraisal interviews and psychometric testing. Together they are so ineffective that according to a delightful piece of research by the University of Catania, companies would be no worse off if they promoted people at random.

So if we are in favour of meritocracies, we should also be in favour of anything that helps us measure merit more accurately.

While the data collected by the new sensors are almost certainly too crude to offer much help now, I see no reason why in time (and probably quite soon) we will not have worked out exactly which behavioural quirks are the key to high (or low) performance, and found a decent, objective way of measuring them.

You could say that monitoring behaviour in offices would kill trust and spontaneity, making robots of us all.