Categories
Bike Share London

Lockdown and Bikeshare

I earlier this week spoke at a Cycling@Tea-Time seminar, on the impact of lockdown on bikeshare, looking at London, the UK, and the world in general. The talk was based on some very preliminary crunching through some CDRC datasets to see how usage has changed, both in volume and time-of-day, for how people are using bikeshare systems.

I also offered some thoughts on bikeshare’s role in a post-lockdown world, where social distancing concerns about public transport may result in a spike in bikeshare usage but also more congestion.

The talk also paid tribute to Russell Meddin, the “godfather” of bikeshare, who sadly passed away last month.

I met up with Russell regularly over the last 10 years to talk bikeshare, and we would typically spend hours over a hot chocolate, catching up on what was happening in the industry, in the USA, the UK and elsewhere. Russell also was the driving force behind many of the changes to Bike Share Map I made over the years. He will be greatly missed.

Amongst many other societal contributions, Russell spent the last 11 years curating the Bike-Sharing World Map, a huge Google Maps site showing the latest news and status of around 2100 active bikeshare systems around the world, along with notes on the 400 proposed and 500 closed systems.

There is no other resource that comprehensively maps bikeshare throughout the world, including my own Bike Share Map that only shows the larger systems with live data. I am sure I am not alone in wanting this resource to live on and continue to be the definitive source of bikeshare’s world “footprint” and would like to explore some ideas about this could happen.

My talk only touched about the impact of lockdown and there is much data that needs to be crunched so I am hoping to spend further time on looking at this shortly.

My presentation:

Categories
London OpenLayers

London’s Poverty Profile 2020

Trust for London (TFL), a charity and themselves a major funder of charitable projects in London to address poverty and inequality, has this week launched the London Poverty Profile (LPP) 2020. There is an updated data-driven website with over 100 different indicators of poverty and inequality, compiled by WPI Economics, along with a PDF report snapshotting the indicators as at early 2020.

With the ongoing Covid-19 pandemic and resulting lockdown likely to cause a significant impact on London’s social economics and community wellbeing throughout this year and going forward, the LPP 2020, which was compiled with pre-Covid-19 data, acts as an important baseline, looking at London’s poverty and inequality profile towards the beginning of the year.

As one of the world’s most international and wealthy cities, it is easy to overlook that London also has areas of extreme poverty and deprivation. The luxury apartments of Knightsbridge and Chelsea are often in the headlines but less obvious are the endemic poverty that has persisted in areas such as much of Newham borough in east London, parts of Tower Hamlets close to the glittering lights of Canary Wharf, or even North Kensington in the west. The recent political focus may have been on “rebalancing the North” (of England) away from London as a whole, but treating London as a single unit of the wealthy South is over-simplistic. The London Poverty Profile acts to ensure that all of London is understood and its challenges, when considered at detail, are not overlooked.

The Consumer Data Research Centre (CDRC)’s London hub has been involved with the LPP 2020 and will continue to work with Trust for London going forward. Our role has been two-fold. First of all, I was seconded to Trust for London periodically over the last year to overhaul the mapping system that appears on the LPP webpages. Previously using a heavily simplified representation of London boroughs, it has now been rewritten to use OpenLayers 6 (in Javascript ES6 form) which is integrated with the Content Management System used to publish the data and indicators by WPI and TFL. Secondly, CDRC will be contributing and mapping “experimental” datasets, from time to time. These will utilise CDRC’s own datasets and its ability to cross-tabulate datasets from other source, open and non-open, to provide further innovative insight into spatial aspects of poverty and inequality across the capital’s 9 million population.

Geographies that can now be used extend beyond the London boroughs, to include LSOAs, MSOAs and (shortly) Wards. This allows more detailed maps. Poverty does not stop at London borough boundaries (although there are a number of cases where there is a big change, for example Redbridge to Waltham Forest), and some boroughs, such as Haringey, are well known for having a considerable east-west split, with a major railway line acting as a physical and socioeconomic split between wealthy Highgate and Muswell Hill to the west, and poorer Wood Green and Tottenham to the east.

Sometimes, other political boundaries do show a step-change in deprivation, as seen here between Ilford South and Barking constituencies (which is also a Redbridge/Barking & Dagenham borough boundary):

In addition, the maps use a selection of ColorBrewer colour ramps to ensure that spatial trends in the datasets are easily seen. ColorBrewer is widely used in the digital cartography field to ensure visually fair and effective use of colour in showing quantitative data.

All maps include a postcode search widget, and ones showing data at a final resolution than London boroughs include a toggle between borough outlines and Westminster political constituencies. Maps are zoomable and pannable, and PDFs and images can be quickly produced.

For launch, the new maps on London Poverty Profile include:

In addition, a number of existing maps on the LPP have been brought over to the new system, and other datasets, typically those split by borough and with some slight of spatial autocorrelation, will also gain maps in due course.

We hope to introduce additional experimental datasets, and corresponding maps, to the London Poverty Profile, on an approximately monthly basis this summer. Possible examples, based on current maps on CDRC Maps, include mapping on access to broadband, rate of household composition turnover, and consumer vulnerability to marketing practises.

Understanding the spatial characteristics of London’s poverty, inequality and other social challenges, is vital, and our hope is that these maps will help inform and better navigate the data available.

Categories
Bike Share Conferences

Walking & Cycling Innovations

I was invited by organiser Landor LINKS to speak at the Walking and Cycling Conference which took place in Manchester last month. The conference included a good focus on bikeshare, and it was a good time for the UK-focused bikeshare industry to pause and take stock of a busy 2019. Three UK-focused bikeshare operators – Freebike, Beryl and Nextbike UK – were present, and it was good to chat with the respective teams and find out how the year had gone and their thoughts for the following year.

MicroMAAS and the UK

I presented on “MicroMAAS” data – first defining MicroMAAS as mobility share services that you can pick up (i.e. bikeshare and escootershare) and outlining the different types of bikeshare popularly available:

I then talked about the “why” of open/standardised data in the sector:

and the “where” – Europe is well behind the US here:

I mentioned CDRC’s excellent and huge collection of largely dock-based bikeshare dock data, available through the CDRC Data Service:

The last part of my talk touched on managing such systems, including key players in analytics platforms:

I also outlined and bemoaned and the (little) progress towards fourth generation bikeshare systems where payment is fully integrated into how other transport modes are paid for, rather than being app-siloed. Right now we are in a commercial battle, with providers looking to integrate vertically rather than horizontally – largely due to the weak management of the sector by local authorities here in the UK – who seem happy to take money and less happy to regulate the sector properly and effectively so that MicroMAAS will actually be a net benefit to the wider UK streetscene:

Beryl Update

Of the other talks, I was particularly interested in Beryl’s – especially they included some data on their first half-year of operations. UK bikeshare usage data is still rather sparse so it was good to see these numbers in a public presentation. The London operation is very small – they quickly moved out of Enfield after the system was heavily abused and little used there – and the City of London “square mile” only has limited need for journeys within it:

Slide © Beryl Bikes (from their presentation)

London’s on-street available fleet is typically around 144 (and around 100 currently) rather than the 400 mentioned here. With approx 5 months between launch in July and the early December presentation, this suggests around 30 rides a day or just 0.2 rides/bike/day (as a rule of thumb, for a non-electric system, over 1 is just about OK, over 2 is good and over 3 is really good – for electric you need 2+ due to the extra costs of the bikes and retrieving them to charge). As you can’t really do a point-to-point journey in the City that is longer than a mile and a bit, this would explain the average journey being just over a mile – half that of Bournemouth.

This may improve with their extension to Hackney that is happening now – so far they have moved into Shoreditch and Hoxton in the south of the borough, but in time if they move into parts unserved by Santander Cycles then they become the cheap, manual alternative to Uber’s JUMP here.

However their numbers for Bournemouth and Hereford – the latter helped by a generous public subsidy – are much more positive. Bournemouth launched in mid-June and averages around 300 bikes (although 140 bikes currently) – so 1 t/b/d, and Hereford launched at the end of July, averaging around 160, or 1.3 t/b/d. Bournemouth is suffering from theft though.

JUMP

I’m also hearing good numbers coming out of Uber’s JUMP system in London – so it is possible for commercially-led bikeshare systems to work here in the UK, it just takes a lot of experimentation, effort and investment.

See also Bikesharp, which is my blog exclusively dedicated to the minutiae of the UK bikeshare market.

Categories
Bike Share London

Consolidating Dockless Cycles in London

This is a draft piece of commentary and I will evolve it in response to any feedback and further analysis I am performing.

A bylaw is being drafted between the 32 London borough councils (and the City of London) to introduce a coordinated approach to managing dockless micromobility sharing, such as bikeshare and (should future national legislation permit it) escootershare, across London.

Currently, each council sets its own policy with regards to dockless cycle operators in their area, making running a pan-London system painful for operators, and resulting in a number of inconsistencies. The matter is further complicated by the parking of a bicycle on a pavement not actually being illegal currently, as long as it is not obstructive, and by “red route” roads in London – the larger roads, which are generally managed by Transport for London and not the councils – and which in some cases have good segregated cycle lanes installed by the transport authority which is more focused on getting people travelling efficiently throughout London, rather than entirely within small borough boundaries – some councils tend not to consider than someone would ever want to leave the borough, as evidenced by mandating max/minimum bike numbers on operators who then watch as their users head, like everyone else, in the direction of the City/Westminster/Canary Wharf, in the morning.

At the same time, there are currently 7 operators in central London (3 free-floating, 2 hub-based and 2 dock-based), a mix of bike types (3 electric systems and 4 manual ones) and yet, while some areas have 5 operators, a third of boroughs have none.

The bylaw will ask each council to outline its policy of where parking of dockless bicycles is allowed, the policy then applied consistently to all operators who want to be in that borough. This potentially could result in huge variations – some, like Islington, may be happy to allow parking whereever, as long as basic sensible parking considerations are taken into account. Some may designate only a small number of hubs, perhaps far away from their local commercial centres and bus/rail stations, where they are out of sight and with little impact, but not useful for the great majority of people. And some may take a balanced approach, like the City of London which has designated (and marked) a number of hubs throughout its area.

My personal view is that one size does not fit all, and in fact there are five distinct categories of publically accessible “realm” in London which all need different approaches to how dockless micromobility should be parked on them.

  • For outer London boroughs (Z5-6), with low population density, the designation of hubs is I think vital for a bikesharing service to ever be viable. But these should be recommended rather than mandated. There should not be any specified exclusion areas, instead, users should follow “common sense” principles.
  • For inner London boroughs (Z3-4) where cycling to the centre of London is viable – on a pedelec at least – it is important to allow the operators to position their bikes where they feel they can provide a service that is viable for them. Councils should publish geo-files containing exclusion areas, such as the busiest pavements in their urban centres, while still allowing the parking of free-floating bicycles close enough to them. If an inner borough is very keen on having designated hubs, then they should either exist on an optional basis (like for the outer boroughs) or at the density of the city centre (i.e. with no part of the borough more than a ~400m walk from one). Hubs must be outlined in brightly coloured paint and with a generic caption like “dockless parking”, and ideally with a metal sign to increase visibility. As below, hoops/fences are an alternative.
  • For the city centres (the area covered by Santander Cycles, roughly Z1-2) free-floating will not work – there just isn’t enough pavement space. A high density of hubs should be made available – these should – as a minimum – include the ends of all the existing Santander Cycles docking stations, as these have a good density throughout the city centres and almost always have space at either end for at least 3 or 4 dockless bicycles – parked at right angles to the Santander Cycles. I regularly see them being used in this way already. Other hubs should either be as rectangles taped/painted on the ground, or designated fences, cycle hoops and other structures to which the bicycles can be secured (using cable locks present in the JUMP system – other operators would need to adapt their bikes to have cable locks).
  • Royal parks (and other urban parks) should adopt the city centres approach of having mandated docking areas within each park (although not at city centre density) – a suitable number around the perimeter of each park, but also one at all their park car parks. If people can drive into a Royal Park car park, why shouldn’t they also be allowed to start or finish a bicycle journey there?
  • Canal towpaths (and the Thames path) are generally linear and cramped, and the adjacent water is always a tempting target for vandals, so bicycles should continue to be not be parked on these – although allowed to move along them. Generally, the nearest designated hub will only be a short distance away from the tow path. Similarly for railway stations and markets.
Building densityDocking station/hub density
Suburbia, Urban parksHubs, ~ max 500m walk.
Inner CityDockless.
Some hubs in retail/office areas.
City centres,
Railway stations
Existing docks (where present)
plus “infill” hubs, max ~300m walk.
Canal towpaths/
river walks/
highwalks
Not allowed.

Other thoughts:

  • Operators should pay a fully refundable deposit for each bike, to the body managing the bylaw, which should be refunded when the bike is withdrawn from operation. This would ensure that operators, to the best of their abilities, retrieve broken bikes and remove them from circulation. If an operation folded, then the deposit can be used by the councils to remove the bikes themselves.
  • Operators should not be charged by the councils (i.e. should not have to pay for permits to operate), except on a cost-incurred basis.
  • Operators must publish the live locations of their available bicycles (when they are not in active use or transport), regardless of whether they are in a hub or not, on a timely basis (e.g. updating every minute) as open data. A suggested specification would be GBFS.
  • Councils must publish the spot locations, names, geographical extents and capacities of their hubs (where designated) and their exclusion zones, as open data. A suggested specification would be GeoJSON. These should be published to a central location, e.g. the GLA Data Store, and kept up to date.
  • A standard way of reporting mis-parked bikes should be adopted, such as FixMyStreet.
  • Councils should have the right to fine operators for mis-parked bikes but only if they have been demonstrably not made an effort to retrieve a bike after it is reported to them by the council, that it is a legitimate report, and after a reasonable amount of time (at least 12 hours from the report being passed on), and on a per-issue basis. The level of the fine should be two-tier based on whether the bike is in an obstructive position or just in an excluded area.
  • Boroughs should fund the cost of marking hubs.
  • Hubs can be on both streets and pavements – if the former, they should be protected from errant car tyres by using “armadillos” or similar equipment.
  • If operators want to fund hubs, that’s OK, but there should not be operator-branded hubs.

Finally – London’s bikeshare operators are actually, generally, providing a good service now. We aren’t seeing the huge levels of complaints about poor parking which were seen when the larger Mobike, ofo and oBike operations were running. JUMP are reporting great usage rates, and the smaller hub-based operators (Freebike and Beryl) have tightly managed fleets. Even Mobike’s much reduced fleet seems to be operating in a less intrusive way, and although data on Lime is difficult to get, it too appears to be operating effectively, in terms of rides vs complaints.

Categories
Bike Share Data Graphics London

Use vs Theft: Risks and Rewards for Dockless Bike Operations in London

Cycle use rates/1000 pax (green) and theft rates/1000 pax (red) in London boroughs. Yellow dots show individual cycle thefts in 2018-9. The green/red borough colour compares the theft rate with the usage rate. Populations are daytime and nighttime, averaged.

When running a fleet of dockless bikeshare bikes, one of the potentially most costly problems is theft of the bicycles. They aren’t attached to anything if they are dockless, even if they are in a marked “hub”, and, even if the bikes are typically heavier than a personal bike, they can still be easy targets for theft. There are six operators in central London currently and each of these operators has to consider whether it is worthwhile operating in a particular borough – whether the profit to be made from legimitate hires outweights the costs involved in replacing stolen bicycles.

With the news earlier this month that Beryl is suspending operations in Enfield due to vandalism after just three months of operation, and following Urbo’s similarly rapid arrival to and departure from the borough (and indeed all of the UK) last year, I’ve done a simple analysis of the risk/reward of operating in different London boroughs. This analysis is an alternative approach to a previous model that looked specifically at general vandalism rates and usage rates, because it looks at the daytime as well as nighttime populations.

I’ve used the Census 2011 Travel to Work counts, comparing the full 16-74 population with that that travels to work mainly by bicycle, looking at both the Workplace populations (i.e. daytime/evening) and the Residential populations (i.e. nighttime/weekends). A simple approximation of the populations is achieved by equally weighting both figures. This means that Croydon’s average population more than halves its nighttime population during the day, while Westminster’s triples. I also only looked at bikes being used to regularly travel to work, as these are the ones that are most likely on the streets, and therefore much more vulnerable to theft.

I also use the Police data statistics on cycle theft, for 2018-9, looking across the Metropolitan Police, City of London Police and British Transport Police force data. I only considered bicycle theft rather than vandalism, as the latter is not broken down by object type, and I believe that general bicycle theft is a good proxy for vandalism and theft of dockless bicycles – with vandalism often occurring as a result of attempted theft. Dockless bicycles are probably not numerous enough in London yet (there are maybe around 3000 available) compared with the ~200000+ private bicycles that are used to commute to work daily with many left in public parking facilities, albeit almost always chained to an immoveable object.

I was keen to not map areas of high cycle theft or use – but rather map one compared to the other. Some places see very little cycle use – the low green numbers – e.g. Harrow and Havering. But they still see some cycle theft – the red numbers – and so the average number of thefts per bicycle is therefore high. On the other hand, Westminster, the City and Islington also see high theft rates but these are more than balanced out by very high usage rates. Only in Hackney, does the very high cycle usage rate (84 bikes/1000 people) still suffer from the also very high theft rate (12 bikes/1000 people). In Hackney, you’ll therefore probably suffer a stolen bike every 7 years on average. In Redbridge though, it’s 1 every 4 years – there aren’t very many bikes in the borough at all, but the few that there are often victims of cycle theft.

This is a really rough study – it could be improved by using more recent population/cycle usage data (which is available for residential areas but not work areas), by looking at vandalism as well as cycle theft, and by more carefully modelling the 24-hour population. But it’s good indicator of why Islington, Westminster and the City of London are so popular with operators, despite a high “headline” rate of theft when looking at the raw Police numbers, and why Greenwich, Newham and Kingston have no operators at all, despite plenty of regular cyclists. It is also why boroughs that sit in the middle – Enfield, Croydon, Southwark and Hillingdon – are probably only going to succeed with dock-based approaches, and so likely require council capital funding rather than hoping that dockless operators will be able to run a successful commercial service for making bikes easily available to those that don’t own one or have one handy – which is what bikeshare is.

Data used in this study:

Another view of the same data – here, the numbers are showing the annual theft rate per 1000 bicycles.
Categories
Bike Share Reviews

The State of Mobility: MaaS Consolidation on the Horizon?

Mobility is a complex and important topic in geography, planning and technology. My research only touches on a small part of the field, namely automated micromobility services (aka micro-MaaS?) such as bikeshare and escootershare, so it’s always interesting to see a wider viewpoint.

As such, I was interested when an acquaintance at HERE Mobility, an autonomous part of HERE Technologies (a major location platform provider), mentioned a new report they’ve recently published, the State of Mobility 2019. While there’s a myriad of information sources on mobility, which has evolved rapidly the last few years, with increasing urbanisation and big technology players funding driverless car research, a single document is a helpful read to keep track of what’s going on.

I’ve used the report to frame some of my own observations of the mobility space, as it stands, rather than a simple review of the report. So, to see HERE Mobility’s own take, you’ll need to download the report (signup link above).

Mobility + Cities = MaaS, Right Now

The report is clear that Mobility as a Service (MaaS) is the current driver of mobility research. That is, shared assets are the way of the future. When living in a dense city of the future, only the lucky few will have space for a car, an electric bike (and easy access to a workshop to fix it). Moreover, even if you do, parking it at the other end of your journey will be increasingly tough.

As a personal example (not in the report), 22 Bishopsgate in London, a tower under construction in the City of London, will have a daytime population of 12000, but will have 4 disabled car parking spaces, no regular ones, and 1700 bicycle parking spaces. The other 80-90% will arrive by public transport. Great – but the trains and tubes of London don’t have much in the way of spare space for the extra people at this and other developments. So, MaaS will become increasingly important in such an environment. You need a bike or would prefer a private ride to a meeting? A fleet of cabs or electric bikes are at your service. The system is patchy now, with rival operators of both modes not particularly integrating well – but the options are there and will only become more important – and their integration is crucial for a useful system that serves all. This is an obvious point but also one that HERE Mobility’s business is staked on – as it aims to become an honest broker of MaaS services rather as a provider itself.

The report emphasises that while MaaS technology is going to have to get smarter – we are going to have to get better at utilising the newer ways of moving through the urban environment, too. The report points out too three technology components of MaaS – the backend crunching big data to create a smart fleet and smart usage of it, a mobile app so the user can get information on MaaS options and perform transactions, and the asset itself having technology, being aware of what it is, where it is, and what it is capable of doing – a so called Internet of Things (IoT) platform. For example, your electric bike (aka pedelec) needs to have a good idea of its remaining battery range and whether it is inside an allowed operating area.

Design Globally but Think Locally

Another key point as that the US is not Europe (and neither are Asia, I would add) – and so MaaS solutions in one of these regions is not necessarily going to ride in the other. Another personal example would be bikeshare.

In Europe, we had Asian-origin bikeshares arriving in 2017-8 (Ofo, Mobike and oBike to name but three) but European and Asian cities and city cultures are fundamentally different. European cities tend to not have the huge pavements of Asian cities or huge roads of American and Asian cities, but we do tend to have a problem with vandalism and theft at a level that is less seen certainly in much of Asia. So, a one-size-fits-all bikeshare is not going to work here.

Similarly we are currently having a wave of US-origin bikeshares and escootershares (Bird, JUMP and Lime). Again, narrow pavements may struggle with the physical equipment, although at least technologies have improved to secure assets more effectively.

HERE Mobility’s report uses the example of the fundamental difference of European and US transport networks – with US cities typically being more car-designed, with wider, straighter roads, while European cities have often had a bigger focus on public transport, such as bus lanes or subway networks. If MaaS is going to come in and act as a complement to both types of cities, then it has to be adapted accordingly. Regulatory differences in the regions are also a factor – while the US has been keen to lead on autonomous vehicle research but introducing sections of public roads in some cities an states where such vehicles can be trialled, European cities often restrict cars of all sorts from large parts of their city centres.

The report’s most interesting section disseminates a survey of over 20000 people, around 50% in each of the US and Europe. Within Europe, they split out northern Europe (UK/Scandinavia/Netherlands) from the big continental players (France/Germany/Spain).

The differences between US, northern Europe and southern Europe are noticeable. Unsurprisingly the car dominates as the “primary” transportation mode in all three regions. In Europe a significant minority use public transport, and in continental Europe in particularly, micromobility also makes an appearance, indicating that Germany, France and Spain are ahead of the game not only with respect to the US but with their more northern neighbours. The other modes in the survey: car rental, ride hail and rideshare, have very low usages throughout the surveyed regions. The survey also breaks down by age group across each region and mode type, with the only significant difference being the youngest group (18-24) using public transport a lot more than the other groups – and US 18-24 year-olds using rideshare/micromobility noticeably more:

Transport App Consolidation

As mentioned above, HERE Mobility is aiming to be a “neutral” MaaS marketplace and so the final part of the survey focuses on the current situation on many people’s mobile phones – multiple apps needed for utilising all the transportation options in a city, along with measuring the desire for such a consolidation for service discovery and payment:

The final part of the report summarises the survey looks to the future. The authors note that it’s not all about price and that a more expensive but higher quality commute, if suggested by an app, might win out. Users generally also are not going to keep multiple transportation apps on their phones although they may try them out for a limited amount of time. And finally, private car usage is very much expected to continue to decline. The report sites Whim, a Helsinki based system that integrates all MaaS modes, from multiple providers, into a single app, is resulting in some very positive outcomes after only its first year of operation.

Here in London, and again focusing on the bikeshare services here, we are seeing some limited horiztonal and vertical consolidation, but we are a long way from rival services sharing their provision data. In terms of apps showing multiple services:

  • Uber has its JUMP bike service, and Transport for London (TfL)’s open data public transport information, integrated into its main app.
  • Google has included the TfL public transport data along with TfL’s (open data) bikeshare (through an ITO data brokerage agreement) and Lime bikeshare, and Uber and a couple of other cab and rideshare servies, into its app, although not Uber’s bikeshare. Apple Maps is similar.
  • CityMapper has Mobike, Lime and Santander Cycles bikeshares, but not Uber’s JUMP, along with TfL data but no cabs.
  • TfL’s own journey planner just includes its services.
  • A number of smaller services (e.g. London’s Beryl Bikes) have started to publish location information in open data formats but these are generally below the radar of multi-option aggregators and so have not yet been adopted.
  • Transactions (i.e. payment) involve, in almost all cases, the user getting redirected from their planning app to the providers app, with the notable exception of CityMapper and TfL services – but if you are signed up for their “CityMapper Pass”

So, a long way to go in London and – indeed – the rest of the world.

Thanks to HERE Mobility for sending me a copy of the report.

Categories
Conferences OpenLayers

FOSS4G 2019

Central Bucharest – “Universitate”.

Last week I was at FOSS4G (Free and Open Source Software for Geo) 2019 conference, in Bucharest, Romania. It was the second time I had attended the global conference, the first being back in 2013 in Nottingham. There are also country and region “mini” versions of the conference, including FOSS4G UK which I have also been to a few times. Relatively cheap airfares and hotel costs in Bucharest, along with the conference fee itself being low for early birds, along with the theme focusing on open source geo software that I use heavily (e.g. QGIS and OpenLayers) meant this was an obvious summer conference to go to. As it is the “canonical” conference for the industry, it means that many of the key technologies have core developers attending – and speaking. Hearing insight from the creators – rather than just vendors – is invaluable.

I attended the main conference days on Wednesday, Thursday and Friday. It was a packed event, with 11 simultaneous streams of talks, starting each day at 9am, and with social events in the evenings too. Two smartphone apps for the conference were a must – Attendify was a good interface work out which sessions to go to when. The app is full of annoying quirks, and ironically lacking on the map front, but does have a bookmarking system which was invaluable. Telegram was the other app, as there was a FOSS4G event group chat which was lively and frequently updated. Around a third of the 1000 attendees were on the group chat. Unfortunately I had to find out about both Attendify and Telegram from other attendees – registration didn’t tell me about these. You would have had much less of a conference experience without these apps.

Openlayers 6.

Day 1 – Wednesday – was the best day for me, as it included core developer talks on GDAL, OpenLayers 6 and QGIS.

The GDAL talk included mentions of ndjson (and so ndgeojson), whichI hadn’t heard of before but is being supported in GDAL 3. It also touched on PROJ 6 and TileDB.

The OpenLayers 6 preview gave a good insight into the main optimisations and improvements being made – faster Vector Tiles, high-volume point-based vector rendering and arbitrary HTML elements as part of the map, to name but three. At the 2013 conference, OpenLayers 3 Beta was released – we’ve come a long way.

The QGIS talk previewed some of the work in progress in 3.10 and the next LTS release. QGIS 2 was launched at the 2013 conference – again, we’ve come a long way.

But the biggest surprise for me was one of the first talks – on bikeshare data visualisation, by Oslandia, detailing their python-based web application showing flows. My own Bike Share Map won Best Web Map back at FOSS4G 2013, and since then the industry has evolved a lot, and my map with it. I wasn’t expecting to see much in the way of bikeshare at FOSS4G, it being very much a technology rather than transport conference, so it was a nice surprise.

Bikeshare data clustering using open software.

Day 2 – Thursday – was not quite so unmissable for me, although this may perhaps have been due to the icebreaker event at Bragadiru Palace, following by a long walk back through central Bucharest in the evening heat – stopping off at Caru’ cu Bere, an intricate neo-gothic pub/restaurant that reminded me of the Cittie of Yorke in London.

Anyway I enjoyed the talk on PGRouting although I would have loved to have learnt about the differences between the main routing algos that have recently been added to it.

The best talk (for me) of the whole conference was on this day and was one that I hadn’t even been planning on attending. It was “Analyzing floating car data with clickhouse db, postgres and R”. It only had a teeny bit of R in it (I’m a python person) and was a great example of crunching a big dataset (all major road vehicle speeds and weather conditions grid data) using a specialised database, and visualising effectively.

Car speeds in the Netherlands and the effects of weather.

Day 3 – Friday – kicked off with “What’s new in PostGIS” was another unmissable talk – PostGIS’s importance in the FOSS4G community being acknowledged by it being a plenary rather than parallel-session talk, and so hosted in the 1000-seat National Theatre auditorium. Unfortunately this meant it was on at 9am, and so I opted to watch this one on the excellent, high-quality live stream being broadcast by CCC, from my hotel room, before checking out and going to the remainder of the conference. CCC’s production quality and coverage is perhaps one of the best things of the entire conference.

Other good talks on the final day included an integration of OpenLayers with the decidedly non-open-source Power BI platform, and a demonstration of Martin (a PostGIS-based vector tile server written in Rust).

QGIS on the Road sounded promising, but was a little too contrived (using QGIS to plan a bee-keeping hobby) and also too long – it was a triple-length session unexpectedly without breaks and was more of a tutorial. I was hoping there would be a demonstration of QGIS on mobile devices.

Finally a talk on GNOSIS style sheets – I certainly think any consideration of good cartography is a good thing, but feel there are already excellent ways (e.g. SLD, or Mapnik/CartoCSS) of standardising cartographic style sheets on the web.

Martin, a vector tiles server.

As previously, there were some themes that I would have liked to see more of such as on advanced Mapnik usage. There was also little on Leaflet, which was a surprise. Heavy users of the open source geo-toolstack didn’t have a huge presence – e.g. Mapbox. Indeed, Google and ESRI, two non-open vendors, were more visible. Like back in 2003, there is little if anything on D3. I was also surprised to see little mention of MapShaper or Shapely.

The other thing was that the OSGeo AGM sessions, representatives of the many incubator and other supported projects had around 30 seconds each to introduce their work and progress in the last year. I hadn’t heard of many of these, and ideally, every OSGeo incubator and supported project would have a least one 20 minute talk during the main conference itself, as an audience education. Perhaps something for the future conferences.

Opening Plenary session in the impressive National Theatre.

So overall another excellent, well organised conference with many good talks and also excellent community networking opportunities. The facilities were good (even if the hotel changed the names of some of the rooms after the programme went to press!) and there was something for everyone in the community. I don’t know if I’ll make it to FOSS4G Calgary in 2020 – I probably should start writing some open geo software first – but hopefully I will make it to another FOSS4G before too long.

Lime eScooters-for-hire in Bucharest, with the older bikesharing system behind.

Micro-MaaS in Bucharest

A note on Bucharest’s micro-MaaS options – it currently has a third-gen non-electric bikeshare, L’Velo Urban although it covers very little of the city, and requires potential users to go one of two manned booths during working hours to get a pass to operate it. So hardly user-friendly. I saw a grand total of one person using the system during my entire 3 day stay.

The other option is eScooters – Lime and Wolf-E are both present. Lime is surprisingly expensive – the equivalent of 60p to start a journey and then 12p/minute. Nearly as expensive as London, in a city where food, drink, the metro and taxis are far cheaper. Indeed, it’s probably about double the rate of getting an Uber. Despite that, there were loads of people using Lime – I didn’t see anyone using Wolf-E. So, Lime may be on to something – there are plenty of people who are brave enough to scoot on the roads (which are dominated by traffic bombing along a way that London traffic doesn’t) and happy enough to pay for what seems like an expensive option – perhaps because it is the only fast option that doesn’t get held up in the pretty bad traffic the city has (there are not many bus lanes either).

Bucharest itself was a pleasant city to visit. Once I had got used to the traffic, it was quite nice to walk around, particularly in the evening-time when the worst of the heat has passed, and it still felt safe to walk around. It is a city with a recent history, with much graffiti (including on historic buildings), crumbling pavements with mysterious holes, and an oversupply of administrative buildings, a place where the car is king (some pavements are unwalkable due to parking on them) – but also a busy, bustling place full of interesting cafes and bars.

A building within a building in central Bucharest.
Categories
Bike Share London

Then There Were Eight

Freebike – London’s newest electric bikeshare system.

Two bicycle sharing systems have launched in London in the last fortnight, joining four systems already on the streets of central London and two more on the edge of the capital:

Freebike has launched an electric-assist system based in the City, Islington, Hackney, Camden, Kensington, Chelsea and parts of Lambeth and Wandsworth along the river. Essentially, central London but excluding Westminster and Bankside. There are around 200 bikes in the initial launch, painted flourescent yellow and black.

The system uses virtual docks. You can pause your journey (at a reduced rate) in the operating area, and also in Westminster and Bankside. You can also finish a journey away from a dock, for an additional fee. Hackney doesn’t yet have virtual docks. Freebike’s unique proposition is that you can do short non-electric journeys for it for free, once you have an account and have deposited £1 in it. The bikes are electric-assist, use of this is optional and if you ride under your own pedal power, it is cheaper!

Freebike is an electric version of the Homeport platform, which already runs smaller systems in a number of UK cities including Oxford, Nottingham and Lincoln, as well as in a number of Polish and other European cities.

Beryl Bikes – at the launch in central Enfield. A marked dock is on the left.

The second launch is Beryl Bikes who are now operating in Enfield in north London. They have plans also to launch in the City of London – along with Freebike, they are the two operators that the City of London have approved for using virtual docks within the Square Mile. The bikes are painted turquoise. Their initial fleet is 350 bikes, covering the full borough of Enfield but focused on the west and central parts.

The system is not electric-assist but the bikes do come with solar panels for charging the lights and also the bicycle symbol laser-lights which were invented by Beryl and appear on the larger Santander Cycles system in central London.

One of the marked docking stations in Enfield.

You can only start or finish a journey in one of 50 virtual docks. Notably, these have been marked out on the ground, as rectangles which often (but not always) surround existing bicycle parking hoops. The bays are also coloured turquoise, and can be used for any bicycles, including future virtual dock and dockless systems in the future, although Beryl do have exclusivity with Enfield at the moment. Beryl should be extending into the City of London soon – they are waiting for the virtual docks to be marked on the ground there first. Freebike will also be using these docks.

The careful and considered launch of these two new systems is a contrast to the existing “pure” dockless systems of Lime, Mobike and JUMP which don’t currently designate virtual docks at all (Mobike did briefly, a while back). It will be interesting to see whether “docks” are the future of “dockless” – whether they can provide the balance between cost-effectiveness of not needing the Santander Cycles docks with their associated planning, pavement reconstruction and power requirements, and order of ensuring that the bikes should be available only from well-marked and sufficiently spacious locations.

Along with the six systems mentioned above, ITS operate a very small two-docking-station system using Smoove bikes (a French company who also supply the Velib in Paris) between the two campuses of Kingston University, using pedal-assist to get people up/down Kingston Hill. Only students and staff can join this system. There is also a small nextbike-based system servicing mainly Brunel University and Uxbridge town centre. Unlike Kingston’s, anyone can use this one. It too is dock-based, but has no electric assist. Nextbike supply numerous systems around Europe and Asia, including the forthcoming huge Birmingham system. Confusingly, the Brunel system is also called Santander Cycles, despite being incompatible with the Santander Cycles in central London.

A quick summary of the eight London bikeshare systems currently operating:

NameSantander
Cycles
MobikeLimeJUMPFreebikeBerylKU BikesSantander
Cycles
Launched20102017201820192019201920172019
# Bikes9000160010003502003502040
ColoursRed
+ Navy
OrangeGreen
+ Yellow
Bright
Red
Bright
Yellow
Turqu
-oise
Yellow
+ Black
Red
+ White
PlatformPBSCMobikeLimeSoBiHomeportBerylSmoovenextbike
OperatorSercoMobikeLimeUberFreebikeBerylITSnextbike
Dock
Type
PhysicalNoneNoneNoneVirtual**TapedPhysicalPhysical
Extendable
Bike
Type
PedalPedalElectric
Assist
Electric
Assist
Optional
Electric
Assist
PedalElectric
Assist
Pedal
London
Area
InnerInner,
West
Inner,
NW, SE
InnerInnerNorth
***
KingstonUxbridge
Ride Cost
1×10 min

“Dabbler”
£2£1£2.50£1.60£0 (ped.)
£1 (elect.)
£1.50£1£1
Ride Cost
2×15 min

“Errand”
£2£2£6.50£4.40£1 (ped.)
£4 (elect.)
£3.50£1£2
Ride Cost
1×60 min

“Tourist”
£4*£3£10£7.60£2.50* (p.)
£6 (elect.)
£4£1£2

* Stopping/restarting the journey at intermediate docking stations will reduce this cost.
** Will also used taped docks in at least the City of London, once they are constructed.
*** Additionally launching shortly in the City of London.

Of note, Freebike is the cheapest public system (i.e. discounting the private KU Bikes) for two theoretical fifteen minute journeys by a user without a multiday membership – both in electric assist and full manual pedal mode. Lime is noticeably more expensive than all the others.

Categories
London

House Price Performance Variations in London Over 23 Years

This map shows how different parts of London have over/underperformed with respect to the capital as a whole, with a 1995 baseline. Green areas have increased in price by more than the London median, while pink areas have underperformed, increasing by a smaller percentage from their 1995 baseline price, compared with the rest of London. Because areas are being compared with their own 1995 price, areas already expensive back then will be outperformed by new “hip” parts of the capital.

This animation shows the data across the 23 years, on a quarterly basis: The strongest colours represent a greater than +/-30% performance difference, while white represents a less than 10% variation with London’s median.

In general, dark greens show the areas that have become more fashionable to live in, relatively speaking, and therefore have seen a greater than average house price uplift. There is clearly an inner/outer split, but established “nice” areas in 1995, such as Islington, remain relatively “average”, while their neighbours to the east – Hackney, Tower Hamlets, Walthamstow (the southern half of Waltham Forest) and Stratford (western Newham), have seen significant gains. The Stratford/Newham (2003) and Shoreditch (2007) big increases happened several years before that in central Hackney (around 2013), Walthamstow (around 2015) and Finsbury Park (2018). The southern edge of Hillingdon, blighted by Heathrow expansion plans, has performed poorly, as well as historically expensive non-central areas like Richmond, as luxury city centre living has become more fashionable than the wealthy flight to the suburbs of the 1980s and 1990s.

Key (1.0 = Proportional to the London median change from 1995)

Data from the ONS HPSSA (House Price Statistics for Small Areas) data files, mapped with QGIS and animated with TimeManager.

Categories
CDRC

Mapping House Prices Across Small Areas

I wrote about a new dataset from the ONS, HPSSA (House Prices for Small Statistical Areas), a few years ago. The dataset has continued to be updated quarterly, and more recently, ONS started publishing the data at a more fine-grained spatial resolution, namely LSOAs (Lower Super Output Areas).

LSOAs typically each contain a population of 1000 people, or 400 houses, so, particularly in cities, mapping house price variation by LSOA, provides a good balance of spatial detail and ease of use. You can of course get individual house prices by looking at the Land Registry Price Paid Data, but the ONS HPSSA is a useful shortcut, particularly as it provides a rolling yearly average, so smoothing out variations caused by low transaction volumes in a small area. The ONS HPSSA data covers all of England and Wales.

I’ve therefore published an updated map of median house prices on CDRC Data, to use the latest release of data, which is Q3 2018. I’ve also extended the key, to reflect that, since 2015, more of London is now firmly above the £500k level which was the previous highest band theshold on the map. The resulting map shows a “dark red cloud” of high-priced areas across much of London, Oxford and Cambridge, with only small areas of cheaper properties standing out in bright yellows – Dagenham, Edmonton and Hayes in London, and Orchard Park in Cambridge. Strikingly, many other cities and large towns also show a small red/maroon area, typically an enclave of expensive houses in an otherwise cheaper urban area (shown with yellows and oranges) – e.g. Solihull in Greater Birmingham, Clifton in Bristol, Hale in Greater Manchester and Gosforth in Newcastle.

Remember that these are median values – so 50% of the houses in each small area, that sold between Q3 2017 and Q2018, sold for more than the value shown, and 50% sold for less. Grey areas are where there were not enough house sales in the year, for a median value to be reported. These tend to be in older inner city areas where little public property transactions take place. Examples in London include Stamford Hill, the area around the just-opened Tottenham Hotspur stadium, and the area behind Euston station in central London, which is being extensively redeveloped. Large areas of social housing, where there simply aren’t properties available on the housing market, also often show up as grey, such as the Aylesbury and former Heygate Estates in Southwark.

The colour ramp is the inverse of that used by Dr Cheshire in his book London: The Information Capital, which depicted house prices in the city using a “fire” colour ramp, with cooler reds with more expensive areas burning bright with yellows/whites, while the highest price, “unaffordable” areas were shown as being completely burnt away from the map. By inverting the ramp, my map shows light, welcoming colours for more reasonably priced areas while inflated values are darkened out.