They say "never go back", but it can be rewarding. If it's ten or 20 years later, you see changes that the locals miss. Infrastructure can have its pattern of use changed beyond recognition.

Two recent examples that have struck me are the M3 motorway and Stansted Airport.

There was a time in the late 1980s that the M3's morning inbound peak period was characterised by a succession of single occupancy private cars. Business parks bloomed around the M25 and commuting trips that would have once involved complex in-out journeys by rail were replaced by much faster car journeys. The succession of executive (generally male-driven) cars was interrupted by the occasional milk tanker.

Observing this same travel peak 35 years later, the gender of the drivers is more balanced, but the really striking shift is that a huge proportion of the vehicles heading for the capital are white vans – often multi-occupied as well as carrying kit. And you'll be lucky to see a milk tanker. The male executives are still there, but they're travelling in a completely different traffic mix.

Van use has increased nationally over the last 20 years by an astonishing 70%, while private car use has increased just 14%. The preponderance of vans on the M3 is because skilled tradespeople are being priced out of homes and premises in London. But the earning potential is so much higher in the capital that even lengthy commutes pay for themselves.

Meanwhile the buzz around business parks has faded as the earnings opportunities have moved on to software and applications rather than hardware and business end-users. Commuters who previously eschewed rail find it possible, with wi-fi, to relax or work. And the milk lorries? Foot and mouth and bovine TB in the West Country, the abandonment of the Milk Marketing Board, and increasing imports of dairy products from the EU are all factors.

In short, not the sort of factors that come into the reckoning in any traffic forecast. Or take Stansted. When I last flew out of this airport about 15 years ago, I was joining that excited phalanx of Brits prepared to fly to scarcely heard of places for the thrill of it: for a price of (say) £29, a weekend somewhere different.

Today the myriad Ryanair and Easyjet flights for Brits to holiday destinations have mostly gone. Instead, the airport has become Europe's labour exchange, with flights linking countless locations across the continent (including Ireland) – but not so many to capital cities or international hubs and even fewer to holiday destinations. That's no longer Stansted's role. It's now inbound not outbound.

Interestingly, Stansted's record high level of public transport use for surface access has been retained. But while rail share has stagnated, coach services, especially those running through the night to and from Stratford, have prospered.

A motorway and an airport, each experiencing a huge change in the profile of users and both adding capacity. And these are less obvious examples: consider the wellknown shifts in the use of ports from exporting to importing coal over the last 35 years.

My point is that you wouldn't really have wanted a detailed analysis of the influential factors that have caused these changed patterns of use 15 or 35 years ago. This is the world of unknown unknowns. We need to be a little less obdurate in our search for the perfect traffic forecast, and a little more enamoured of probing changes in already apparent demand patterns: a narrative around the risks, rather than ever more detailed modelling of behavioural responses in a world assumed to be in equilibrium.

This is a plea for a difference in emphasis, not for less rigour. It leads to the question of how differently a piece of infrastructure might be used in practice over time. Network Rail has made a good start along this route by building widely diverging scenarios into its project assessments. But we need to go further. It doesn't mean more costly analysis. We are now in the era of big data: anonymised data on travel patterns by all travel modes with trip purpose breakdowns – great detail captured from mobile phones.

When you look at the quantity of personal travel behaviour, a measure of travel diversity (in time, routing, speed, linkage – even journey purpose) can be deduced for any specific piece of infrastructure. The multiplicity of customers and diversity of journeys themselves serve as an indicator of whether use of a given piece of infrastructure is likely to be resilient in the face of market shifts.

And meanwhile I shall point to the 20 million-plus passengers using HS1 each year, close to the early long term forecasts in total, even if half of them are travelling on domestic trains unforeseen 20 years ago when the now privatised Eurostar first ran.

Reference: Transport Times, April 2015 Issue

To keep abreast of current transport issues subscribe to Transport Times now, following the link below