Tag Archives: transport models

Transport Data – not just for modelling geeks

Firstly, a massive apology for not posting in some time. Life has been a bit full on over the last year or so, so the blog has taken a back seat. My apologies especially go to those comments that have remained unmoderated for 6 months. My bad. ūüė¶

Also, just to let you know that while I will touch on the Streetgeekery stuff every now and then, this blog will take on a more transport technology and futures focus. I will still remain true to its original intent, just less about street lamps and tactile paving. Besides, there are loads of great blogs covering that. Anyhow, on with the post.

Pedestrian Modelling

A pretty pedestrian model. Source: Wikipedia

If you think about the UK’s contribution to the transport world you will no doubt think of the likes of Brunel, George Stephenson, maybe even maritime (the whole Britannia rules the waves thing). But one area where the UK has made an undervalued contribution is transport data. While Detroit may have led the way with four-stage modelling, and¬†Lowry’s model of Metropolis form the basis of modern transport modelling, the early work of the Transport Research Laboratory set the standard for data collection, and standards required of its quality.

Now, this blog has been critical of models in the past (and boy has it been), but they are and are likely to remain a useful tool on which to inform transport decisions. But the data collection game to inform these models, and the data collection requirements of transport professionals, are changing in a digital world.

The sheer amount of devices that can collect data on transport (there are over 83 million mobile phones in the UK alone) but tracking movements isn’t enough. There is an ever-expanding sea of personal information willingly shared out there, with an exploding number of things to collect it, and even more opportunities that it brings. Businesses like transportAPI are finding new ways of manipulating this data, and generating new business models from it.¬†The shift is taking place where simply collecting and modelling transport data isn’t enough. Transport professionals need to understand that data creates value outside of major scheme business cases or informing a purely transport decision.

Let me put it this way, and this is something that may be uncomfortable to many. The best users of transport data are not transport professionals.

That sounds nuts doesn’t it?

Ok, perhaps i’ll clarify a bit. If you want someone to use transport data for a specific transport purpose, to a modeller you must surely go. But there are many more companies using open transport data that, frankly, none of us could have thought of. Want proof? Try CityMapper for one – transport data repackaged into something usable. Or how about how Transport for London’s open data now powers over 200 apps?

Transport data courses need to be so much more than building spreadsheet models and regression analyses. It needs to be something more fundamental – how transport data can create value outside of transport, and our responsibilities as professionals to not just rigorous data quality for our clients but to making this data available for the common good. That means going outside our comfort zone, and facing challenge from outside the industry on how this data is used.

A report from the Transport Systems Catapult¬†(yeah, its a work plug, so sue me) is intended to be the first step on this journey. It’s not the entire story but I’ll say one thing: the days of only collecting data from manual counts, and putting into simple spreadsheet models to form the basis of transport decisions are numbered.

What Dr Beeching told us about Predict and Provide

Dr BeechingIt seems to be a bit of an anniversary year for transport reports this year. First amongst them is of course the one report that no self-respecting professional has not read: Traffic in Towns, or the Buchanan Report – 50 years young this year. A much younger report is celebrating its 10th birthday this year: Making the Connections – for me the most important transport-related report of the last 20 years. Another biggie that is also celebrating its 50th this year (1961 and 1962 must have been big years for transport consultants) is The Reshaping of British Railways, or to use its more common and sinister name, the Beeching report.

It is not the purpose of this post to debate in depth the nature of the work undertaken by Dr Beeching and his team. This has been more than covered elsewhere. This post has been prompted by an article in The Observer at the weekend¬†entitled “How Beeching got it wrong about Britain’s railways.” As you can guess from the title, this article goes on about how with the benefit of hindsight, and with growing traffic on the railway, Dr Beeching’s plans were flawed and have held the railway back. A particularly good quote is from Christian Woolmar:

Transport planners in the 60s simply could not conceive of the idea that a line, once closed, would need to be reopened. Their mindset saw trains as dirty and futureless.

Now we should be fair to Dr Beeching, for as the Zelo Street Blog rightly points out he was brought in by Ernest Marples – the then Minister for Transport – with a specific aim: to make the railways profitable. He was not to be a fortune-teller, nor a professor in transport. His role was simply to turn the finances around.

The context in which Dr Beeching was undertaking his works is of greatest interest to us, and I am sure that it played some role in his thinking for the future of British railways. At the time, car use and ownership was rising, and rising fast, while rail patronage was dropping like a stone. The philosophy in transport at the time was clear: predict growth in car travel and reductions in rail travel, and provide for it by building new roads, and closing lines. All of this is based upon evidence at the time that shows these trends, so why should they not continue?

DfT forecastsUltimately, transport forecasting is the product of the data and understanding of its time. Professor Phil Goodwin produced a very interesting graph in Local Transport Today showing the variance between traffic forecasts as predicted by the National Transport Model and actual traffic levels that resulted. I have no doubt that such predictions were based upon the best evidence of the time, and that research of the time was undertaken to the highest of technical standards. But as the graph shows, forecasts are always doomed to fail.

Our understanding of travel patterns, transport behaviours, and the interactions of transport with society and economy continually evolves through research. Painstaking research has developed our transport knowledge in ways never thought probable. This is a problem for predict and provide, which assumes the current knowledge as an almost absolute truth, and our plans must be based on current travel patterns.

The fundamental truth that we must accept as professionals is that the data on travel behaviours always lags behind said behaviours happening, and our understanding will lag even further behind as the research catches up. This understanding is fundamental to future forecasting as a sense-checking mechanism. To use an example local to me, traffic flow data from my local motorway junction shows no changes in Annual Average Daily Traffic Flows for about 7 years. But there have been major works on the M1 during that time, including a complete junction rebuild at this location, there has been a recession, the number of trips annually per annum has not changed in that time. This sense-check, informed by wider and continuing knowledge and research into travel, stops me from taking the current data at face value. But my understanding of this has lagged behind these events occurring, because the research has too.

So interventions based on predict and provide will always lag behind our understanding of transport and travel. Dr Beeching and the Ministry of Transport at the time suffered from this same dilemma. They knew the current travel trends, and based upon the evidence of the time, and they sought to deliver upon that. They could not have predicted the rail renaissance since the 1990s, changes in freight traffic on the railways, and changing lifestyles.

The impression that you may be getting from this post is that I think that predict and provide is fundamentally rubbish, and that we should not even try guessing the future. Nothing could be further from the truth. Predict and provide has its place, but the lesson to take from Beeching is that it is not the place that we traditionally think of it being.

Predict and provide provides us with a base case scenario – if nothing changed, what would be the implications of this in the future? Would there be more congestion? Would our kids get fatter because they aren’t out cycling? Would trains be utterly rammed to the rafters all day, every day?

The question then is whether or not we feel that this base case is acceptable – economically, socially, environmentally. Assuming the answer is ‘not good’ (lets face it, it usually is not good) we then ask the question of most fundamental importance: “where do we want to be in the future?” From there, we define the objectives of what we want to achieve, and based upon the best available evidence deliver interventions that will achieve those objectives. How we deliver these objectives should continue to be subject to rigorous scrutiny and review as our transport knowledge develops through research. But objectives should always¬†guide your strategy and plans, not just historical or current trends in travel patterns.

That is the main lesson that transport planners should take from Beeching. Beeching was set up with a purpose to make the railways profitable in an age when they were seen as a dying mode of transport. His remit was narrow, and his recommendations were based upon the narrow view that current trends in rail and car use would continue ad infinitum. It is just sad that many of my fellow professionals – even those whose views are similar to my own – still have such a view. We shouldn’t plan for what we think will happen necessarily, we should plan for what we want to happen.