Big Open Wireless Data

This is a summary of a number of issues around Integrated Urban Modelling discussed in a panel session at the Future of Wireless International Conference 2015. This picks up on a number of issues raised by the Integrated Urban Modelling post but has more of a focus on data.

What are the implications for a fully connected world and how will this impact on the urban landscape?

The opportunity is to collect data at a scale, resolution and frequency that was previously impossible. This means it’s possible to see patterns of movement across an entire city across the whole day. The real question though is how best to use this, and what do we actually need to make cities better?

There are two clear applications at the moment – short term management of infrastructure and services, and long term planning based on a better understanding of how cities work and scenario testing changes to these cities.

There are lots of interesting examples of how cities manage access to services (Jun in Spain uses Twitter to manage all municipal services), or infrastructure (in Chicago the direction of part of the i90/94 is reversed to create higher capacity).

Making the data open and maintaining it in the least compressed form will maintain its flexibility, allowing new opportunities to emerge which are shaped by citizens. Rather than discuss these in more detail I’m going to show how it could influence long-term planning.

To use for longer-term issues, it’s important not only to identify existing patterns but also to be able to test how changes to the city impact on these patterns.

Many problems in existing cities are unintended, and partly caused by the approaches to planning that were popular in the late 20th century – zoning land uses into mono-functional areas and prioritising the use of the car (at high speed) over all other movement types.

This has been exacerbated by the disciplines involved in city planning which have developed increasingly sophisticated, but siloed models. It means that a traffic model talks about capacity or congestion, but gives no idea of how it impacts on place.

These siloes are refinforced by the data that goes into the model, and the fact that this data has become very specific to that discipline over a number of years. The result is that models do not talk to each other in the way they could or should, and developing a truly integrated approach where all urban systems are designed as a whole is held back at this level.

As wireless data is relatively new, it hasn’t been through the process of refinement and specialisation in response to a particular discipline (or at least it has been through a different route to the normal datasets), which could mean that it has the potential to reach across disciplinary siloes and form part of a common dataset shared by disciplines and departments.

There are a whole set of privacy issues that also need to be resolved, and this may be where the role of space comes in as a modelling infrastructure; rather than keeping track of individual people, perhaps the answer is to aggregate total numbers of people (or phones) to a space.

This also requires the development of models or platforms that are able to coordinate these datasets and provide outputs which are useful to planning, and which relate to the things a city can control.

The impact is likely to be very political for many reasons, including questions over whether administrative boundaries actually reflect the characteristics of a city, and also in that these datasets could provide a very clear means to monitor the perfomance of decisions made by the city.

In terms of the application of this data to modelling; it will have the technological, commercial and/or political structures embedded within it as any dataset does, however because of the scale and frequency this data is collected, it may help reduce the sensitivity of models to these issues. However, better data in transferable formats, means that many disciplines can share the same models. This means that many of the problems of unintegrated planning can be avoided, which means that you should live in a city which also physically better connects people to each other and to places. Land use, density and public transport should be better integrated without losing the potential for unintended interaction.

What will cities be like in 5-20 years time?

The way that data will impact cities will be in two phases – the first affecting “softer” issues, management of systems, delivery of services etc, the second affecting the physical structure of cities – where the big things go that take a long time to build and to change. Its arguable that these are both happening now, but that it will take a few years before we see the impact.

In 5 – 20 years time, I expect softer side to be at a peak while the hard systems will be in the process of being implemented.

Many of these softer systems are emerging now in two broad categories: the development of apps (on the back of open datasets published by cities) designed to improve day-to-day life by and for people who live in the city, and the commercial use of this data for advertising products.

Maybe your phone alarm goes off 5 minutes early because it knows the tube line you use to work is busy and it recommends a different route. Maybe it tells you, if you still go shopping rather than order through the internet, that what you wanted to buy is in stock in a particular shop and so you choose to go there instead.

The big con of this is that from a citizen point you might lose the spontaneity of things. There are two responses to this; first that people don’t have to follow what their device tells them.

The second is more complex and affects the way that land uses migrate to particular parts of a city over time – if a shop doesn’t have enough people walking past it goes out of business (see Bill Hillier’s theory of the movement economy) – so will the way that information on these uses is filtered to customers start to influence this and cause shops to go out of business because they are not recommended enough by users (or by paid advertisement)?

It’s possible, but cities will still be physical things that require people to move through them. There will still be places which are naturally easier to get to and which will benefit from their location. In some ways it’s like the department store model; people still know about department stores but it doesn’t stop the shop next to it from working.

For the professions involved in planning and managing cities this will be a time of transition. New practices and professions will emerge based on the availability of data, and the potential implications for how it could change the way planners, urban designers, architects, engineers, citizens, all work (together). It will result in a period where many ideas are developed and tested, some will work successfully, some wont. It will require not only a transition in the way that professions work, but also in the administrative organisation of cities – collaboration between departments and cross budget spending.

It will also present requirement to find new funding models – who pays for data, the infrastructure to collect it, and the models to understand it? What new funding/income opportunities might this data present, and how can it be used for the public good? interestingly, because funding at the city level is stretched at the moment, many of the applications being developed are happening from the bottom up with local developers making use of open data and innovating, rather than large (or even multi-national) companies providing the ultimate but costly city management tool.

What will cities be like in 50 years time?

I’m going to cheat and give two scenarios for 50 years time (these are both on the assumption that humans haven’t become voluntarily extinct to occupy a data network).

Scenario one: People will be unhappier, unhealthier and poorer

Unhappier because they are under constant surveillance.

This is carried out by the city, to control their movement and to try to prevent civic uprising – this is done by monitoring social media and movement, and providing false information on which parts of the city to avoid/use.

Unhappy because their day to day activity is recorded and used as an endless opportunity to market product and services specifically tailored to them every time they turn a street corner or pass a bus stop or advert.

Unhappy because the constant requirement for data driven solutions sucks energy supplies resulting in blackouts and power shortages.

They will be unhealthier, because self-driving cars will have become an easy answer to everything.

Public investment in public transport networks is not needed because self driving cars will make more efficient, cleaner, use of the existing road networks, on the basis of clever way finding and coordinated management of the entire system. In combination with this convenience, the routes used by these cars will be so complex that people will know their own cities less well, meaning that they stop walking.

Poorer because the jobs they do will have been automated and replaced by software.

Scenario two: People will be healthier, happier, and wealthier.

Healthier, because people and cities will still be physical things that we’re required to move around, however they will be planned in a more integrated way that makes active and public transport convenient and a normal part of day to day life.

Key land uses – such as concentrations of employment, shops, social facilities etc will be in places that are easily accessible to the populations they support. Land uses will be planned in relation to an understanding of how people naturally move through the street network, and how this creates opportunities for interaction. This will create spaces which are vibrant, and which will create opportunities for chance interaction – these areas have also been found to be the parts of cities which create more opportunities for knowledge sharing and innovation.

Happier, because transport, infrastructure, land use, open space and housing systems can be planned in an integrated way, that reduces the likelihood of pedestrian severance which isolates residents from each other, their community and wider services.

Happier because cities will be easier to move through – management of networks will happen in real time, reducing congestion and time required for travel.

Wealthier, because public finances can be spent more efficiently leading to reductions in council tax, because the city will create places that support street based economies (such as high streets), encourage innovation through digital access to tools, training and data to develop ideas, but also physically developing the characteristics of place that support knowledge based economies.

How can we get to scenario two? As mentioned earlier there are many unintended consequences of design, however if these can be reduced through modelling the question is why these unintended consequences keep happening and how do we change this? By integrating the tools that disciplines use to understand the impact of urban systems on each other, and most importantly, on people.

How do we get there – by starting to define data standards that allow people-focussed outcomes to be delivered, and which are collected for their purpose rather than adapted from other data that happens to be available. This of course raises questions about whether, for example, mobile phone data can be used to measure movement or whether a unique set of sensor are required.

Models will not provide the answer in themselves, but if a model is well made to answer key questions, it can help people to make better decisions. Access to data helps understand cities, improve models, introduces a way to monitor the perfomance of change over the long-term, and introduces more accountability around planning.