My biggest concern about some previous books on infographics (and much of what is available online) relates to their uncritical promotion as brilliant ways of displaying factual information based on complex data. I was therefore pleased to see that in her introduction, Sandra Rendgen, stresses the problems with “Suspicious Data” stating that:

“…general popularisation (of infographics) brings with it a level of denigration, and content-related weaknesses are frequently found in graphic representations”.   

After reading this I was reassured that there was a strong editorial policy in terms of the graphics selected and their associated commentary. This shows and, despite the many pages to be filled, that vast majority of what is featured can be held up as best practice. The essays and associated images at the beginning offer good historical context (data visualisation is nothing new) and the location section alone would make for a good book for cartographers (sad not to see an Swiss mountain maps in there though). I also found the Category section (especially the work of Stefanie Posavec, below) particularly interesting by showcasing a range of examples of visualising more qualitative data.

Everyone who comes by my office has opened the book at least once and been tempted to buy it. For £29 on Amazon (RRP £44) I think you would be mad not to. For those producing data visualisations the book provides some great inspiration for future projects, whilst those who simply enjoy looking at them will not be disappointed.

TfL have produced a nice range of maps to promote walking instead of public transport (especially during the Olympics).  The maps use different cartography to the other walking maps TfL have produced,  and feature a nice range of pseudo-3D buildings and key landmarks. There are also 4 concentric circles that give an approximate walking time up to 25 minutes away. These have been centred over the “hotspot station” the map was produced for- in this case Fenchurch Street.

The maps were being handed out to hardened city commuters who I expect do most of their onward journey by foot anyway. Hopefully TfL will continue to distribute these maps right up to and throughout the games. I can’t seem to find an online version of these so keep an eye out for them at your local “hotspot” station (of which there are many).

New York City last week released a preliminary map showing the proposed sites for the launch of its bike sharing scheme, now named Citi Bike (with Citigroup being the lead sponsor along with Mastercard).

Citigroup’s sponsorship is crucial for the scheme, which has promised no public subsidy on at least operating costs, and is a rather convenient sponsor in terms of its name. In several other cities around the world, their bike share schemes are known as City Bikes, such as Stockholm and Vienna, so Citi Bike has a good chance of becoming the “on the street” name for the scheme, unlike the unwieldy “Barclays Cycle Hire” name we have here in London – most people here know them as the snappier, if politically incorrect “Boris Bikes”.

NYC’s scheme is clearly influenced by London’s – its of a similar size, it has a big sponsor from financial services and a mayor fully behind it, and a Boris Bike from London even appears on the front cover of the NYC DOT presentation to communities. The technology used is the same and the bikes are also the same design.

I’ve extracted the data from the official map and added it to my own set of maps for 50+ bike share cities across the world allowing for a direct comparison between the scheme and the existing ones. The initial map has 413 stands – the districts either side of Central Park are missing, as is central Brooklyn, as these areas are still undergoing consultation and will gain coverage next year. The scheme should be opening this summer and is then due to expand by 50% by Autumn 2013.

The stand sizes and descriptions are also from the official map, and I’ve simulated the empty/full status of each stand, based on the distance from Wall Street and random perturbation. This results in just under 7000 bikes, based on a roughly 1:1 empty to occupied stand ratio, which is fairly standard around the world.

New York vs London

* Announced figure. Actual figure may be less due to bikes in maintenance and temporary storage. London’s equivalent figure was 6000 bikes. ** Zone 1 only. Cost higher if travelling to Zone 2 (which has bike share bikes in it). Cost lower if only in Zone 2.

What stands out for me, when comparing New York‘s and London‘s bike share schemes, which are roughly similar in terms of number of bikes and stands, is that NY’s footprint is similar in size to London’s at launch , but with many more bikes, and the scheme is accordingly more dense. Certainly, New York will have the critical mass of stand locations, so allow the scheme to work efficiently – you’ll never have to travel very far, if your destination stand is full, to find another one.

The other thing that strikes me is that all the stands are quite big – very few of them have less than 20 docks. The biggest, on Pershing Square (by Grand Central Station) has 128 docks – this is ever so slightly larger than our own “superdock” at Waterloo Station and presumably designed with a similar purpose of satisfying the commuter “tide”. The other big commuter station in NYC, Penn Station, has three large docks surrounding it. The coverage is also fairly uniform, my only surprise is that there are only two docks in Battery City, which is surely full of people likely to use the scheme – or perhaps they just walk to work? Also there are none in Central Park – although perhaps these will be included in the Upper East/Upper West areas for next year’s expansion?

One big difference is the fee structure – at $10 a day but only $95 a year, this suggests that tourists and public-transport-based commuters are the target users, rather than local residents and errand users. This is a pity – the latter group tends have more heterogenous usage flows and help “mix” the scheme up and redistribute it organically, requiring less redistribution of bikes trucks by the operator.

$10 is four times more than the cost of the New York subway ($2.50/trip with Metrocard) so you would need to do at least four journeys a day to save money. In London, our tube in Zone 1 is £2 per journey with Oystercard) or £1 a day on the Boris Bikes. So end up often using the latter simply on cost, even for one journey. The over-30-minute journey extra cost is also significantly more – $4 compared with £1 here. Subscribers get 45 minutes free rather than 30 minutes. This gives those commuters a chance to travel further in the busy rush hour – although surely this increase the redistribution challenge even further.

NYC’s CitiBikes are thinking big, and the design of the scheme suggests that it is expected to be wildly successful at launch. Hopefully this will prove to be the case!

You can see my map here.

Photo credit: Edward Reed / NYC Mayor’s Office

Jason Silva’s wonderful demonstration of patterns and order across many evolving spatial systems is what our science of cities is all about: flows, interactions, dynamics, self-similiarity, hierarchy and of course fractals. Take a look. These patterns swirl around all our systems from ourselves to that which we create.

The schedule of events for the day was as follows:

12:45 Arsenal played Norwich at the Emirates, attendance: 60,092

17:15 FA Cup Final between Liverpool and Chelsea at Wembley, attendance: 89,102

Evening, London Prepares Event at the Olympic Park, attendance: 40,000

5 May 2012 16:30 BST (45 mins before kickoff). Map shows tube locations taken from the TfL Trackernet API (link to raw data below)

The image above shows the positions of tube trains 45 minutes before the Cup Final kick off. Wembley stadium is located half way between the “y” of Wembley and the two tube lines above it, which is the location of the closest station to the ground, “Wembley Park” on the Metropolitan (purple) and Jubilee (grey) lines. It’s interesting to note the obvious gap in the service on the Bakerloo line (brown) which serves “North Wembley” and “Wembley Central” to the south (where the word “Wembley” cuts the brown line). We can look at the tube status messages from TfL for this time period and see that there are planned closures as follows:

District line (green): Turnham Green to Ealing Broadway

Northern Line (black): Camden Town to Mill Hill East and High Barnet

Piccadilly Line (dark blue): Acton Town to Uxbridge

These can be seen as sections on the map where there is an obvious lack of trains (open the KML links below for the original data containing station names). The significance of this is that any Chelsea supporters living around Turnham Green are going to get pushed towards Paddington to go North. Liverpool fans are likely to be coming from Euston.

If we move on to 20:30 after the Cup Final has finished and as the later events at the Olympic Park are starting, we can see the situation around Stratford (centre of map).

National Rail and Tube trains around Stratford for 20:30 (link to raw data below)

The National Rail trains show as blue, where the service is on time, red, where it is late, and white where the timetable shows there should be a service, but we can’t verify its location. Due to the differences in how National Rail services work, it is a completely different type of data to the Tube. For National Rail we can only look at the departure boards for stations and use the timetable to match up services. There is only one late train for this time period, coloured red and hiding in the top left corner. This highlights the differences in the type of data as it takes several minutes to query enough data from National Rail to make the map, during which time the trains move around, causing the uncertainty in the data.

This is still a work in progress and requires a much more rigorous analysis, but you can see delays occurring around Wembley just before and after the match, plus some services heading for Stratford running a couple of minutes late in the evening. I’ve not got any information on the National Rail closure affecting services back to Liverpool in the evening, but it doesn’t look as if they were any really major problems.

As this was the first attempt at collecting a comprehensive set of data for a single day, it didn’t go completely to plan. There are questions about how you cope with the uncertainties in the National Rail data and how you compare it with the Trackernet information. The DLR and Overground are missing, as are the buses and it’s not clear how to use the TfL tube status information. We also don’t know anything about the commuters on the network, so can only guess at where all their journeys begin and what route they take. What is also needed is baseline data on what a normal Saturday should look like, which will give us the ability to pull anything abnormal out of the data.

Ultimately, the reason behind doing this is to provide a real-time snapshot of London’s transport network and how it behaves over the course of a day. For this we need to establish an automatic method of detecting and highlighting problems which is proving difficult at the moment. Then we can look at how a problem on one line has a knock-on effect on another.

The image below shows an animation of all tube trains for the 16 April 2012 from 8am to 8pm [link to movie]:

Links to data used in this post:

Tube Network KML

Trackernet 16:30 KML

National Rail 20:24 KML

Trackernet 20:30 KML

MapTube Map of Realtime Tube Locations

I was at the third WhereCampEU “unconference” which took place in Amsterdam over the last weekend of April, following previous editions in London and Berlin which I was also at. The meeting was an ideal opportunity for me to feature CityDashboard which I unveiled at the CASA Smart Cities conference a week before, and to show a couple of the items that were popular at the exhibition that accompanied Smart Cities – namely the London Data Table and PigeonSim.

Amsterdam proved to be a challenging city (financially) to visit for the conference, as it was the weekend before Queen’s Day – which is essentially a massive party throughout central Amsterdam, resulting in expensive transport to get there and all the central hotels being booked up or extremely pricy. So it was that I ended up on the outskirts of the city, overlooking a motorway, although this did mean I got to use the very fast and efficient metro service into town each day. Pre-conference drinks were held upstairs in De Waag, the oldest non-religious building in Amsterdam and a fantastically atmospheric venue. The conference venue was a short walk from here.

To get to Amsterdam I took the Eurostar to Brussels, spent an hour and a half cycling around the city on one of the Villo bike-share bikes, and then got another high-speed train to Amsterdam. A nice way to see the countryside, but it did take six hours in total. My return was a 40-minute flight.

Unconferences have no set speaker schedule, but instead participants put a post-it note with their talk title on a grid of times and rooms, and everyone looks at the grid to determine what to go to next. The plan had been to present early on the Saturday and then just relax and enjoy the rest of the meeting, but the Saturday grid was very quickly full, and it wasn’t until Sunday lunchtime that I was able to squeeze in my talk. Although 26 minutes of my 30 minute slot was spent on CityDashboard, most of the tweeted photos were of PigeonSim (that I squeezed in the last four minutes) and my attempts at demonstrating the flying gestures…

There was as usual a wide range of geo and tech talks, one of the most unusual being a psychogeography session with Tim Waters – this unexpectedly involved a practical where we went out in groups and followed and observed pedestrians going about their business (an initial “meta” idea to follow the followers having been vetoed by Tim). I also enjoyed Jeremy Morley’s update on the OSM-GB project at Nottingham to quantify the quality of OpenStreetMap in the UK, and Peter Miller’s peek at a 2.5D rendering of OSM data. Peter also showed behind the scenes of ITO Map’s map layer scripts, these produce simple overlays highlighting particular OpenStreetMap content – these were the inspiration for similar functionality I incorporated into GEMMA. Finally, a short Geo-yoga (mimicing the shapes of countries) session was certainly an eye-opener. Parallel sessions meant I missed some more interesting talks, including one from Google on why Google can work with OSM.

Thanks to all the organisers for putting on another excellent, and free, WhereCampEU!

Blood is everywhere when it comes to describing cities. We have arterial roads, pulsing transport flows, and cities with different metabolisms. Thanks to great new datasets and visualisation software the analogy of cities with pulsing flows is being increasingly utilised for explanatory mapping. For example the work of UCL CASA’s Jon Reades above depicts the London Underground network as a set of arteries that thicken as passenger volumes passing through the network increase, whilst Pedro Miguel Cruz has taken it one step further to depict Lisbon’s roads as “blood vessels” (complete with their own clots).

I think these visualisations offer a neat conceptualisation of city flows provided the metaphor isn’t stretched too far. As the slime mould (below) shows us, we have a lot to learn from comparing our cities to biological processes- just so long as it doesn’t get too gory!