Back in July, I wrote about transport planning for places, rather than individual modes and ‘networks’. Last month, I wrote about transport planning to accommodate the needs of people, rather than the temptations of technology. Last week, I spoke about both at the South West England regional conference for transport planning. Planning, including transport planning, is by definition about looking towards the future and how we create better places than we have now that improve the quality of life for the people in those places.

Yet in preparing my presentation for last week, and in listening to some of the other presentations, I realised that transport modelling, forecasting, and thus planning have yet another loadstone to cast off before they can ‘help shape a better world’, another challenge besides remembering that the best transport planning invisibly serves people and places. And that weight is the weight of averages.

As a methodology for representing individual behaviour, the average, the ‘usual’, falls woefully short. It ignores the steps people may take to be sustainable or exercise more unless they do so more than half the time being measured. It glosses over the people who do not have the same destinations to access on a daily basis. It downplays the regular, but infrequent patterns of linked trips to visit family or participate in other activities that induce diversionary routes once a week or once a month. It gives no thought to how some people may react to increased risk, delay, or disruption due to severe weather, planned events, unplanned incidents, scheduled repair works, or even terrorist threats.

To plan for local contexts, the average assumptions about how people travel to, from, and within areas of particular land uses can easily miss the diversity of options, variety of economic drivers, and cultural preferences in different places. If most traffic and transport models, whether to assess the impacts of new developments or to inform investment decisions with a cost-benefit ratio, are based upon data collected on average dates for an average population and average land uses, it is no wonder that transport planners are still living in a ‘predict and provide’ paradigm. Nor is it surprising that those predictions often turn out to be wrong.

Way back in March, I wrote about Visions of the future of transport and society developed through scenario-planning techniques. I’ve read academic articles advocating scenario planning in order to address the uncertainties we face. But the key to scenario-planning is not only to think about how people behave and how places might take shape, but also to consider a spectrum of possibilities. A spectrum that encompasses extremes, which in turn do allow for hybrid possibilities, but not averages.

This is where big data and new technologies and ‘smart’ infrastructure can help. Algorithms might still regress data back to averages, but that data, those sensors, the digital trail we all leave in our wake like high-tech breadcrumbs , can also give us a much better understanding of extremes than we’ve ever had before. No longer dependent upon snapshots or cross-sections, planners can take a long view and find the patterns of flexibility that better represent the lives we all lead. Instead of predict and provide, let’s propose and future-proof. Because the future is unlikely to be any more ‘average’ than the present.