New Transport Horizons or Mobility Spam?

Dockless bikes show us one vision of a mobility future. What can we learn?

Photo by Jon Russell / Flickr (CC BY 2.0)

Recently, city officials from London to Manchester to Amsterdam and Melbourne have been wrestling with the appearance of Singaporean oBike and similar bike-sharing schemes in their streets. These dockless variants of the public, pay-by-use bike models being launched in major cities around the world allow users to pick up, pay for, then leave a bike anywhere within an operating city, with no organized storage system per se, just free range. As seamless as this might sound in theory, in practice it’s causing headaches that may be yet another signal of a complicated mobility future that’s emerging as societies transition to new mobility models. New public two-wheeled platforms, like many complex systems, carry cultural values, and those carried in some of the latest bike systems speak to what we may experience in an autonomous four-wheeled future.

The fast-entry strategy of oBike and its competitors, as well as the convenience of a “leave them anywhere” model are problematic because they a) inject a new, proprietary appliance into already complicated, clogged urban environments, and b) rely on a particular model of orderly social norm to operate reasonably. These characteristics have the potential to become a more common strategy in coming years, as we see other mobility systems, driven by opaque business models for manual, autonomous, passenger- or cargo-bearing vehicles, getting dropped into our local landscapes.

Convenient carriage, or mobility spam?

These early entrants into dockless bike-sharing programs are plunking new quantities of metal, rubber and plastic at volume into city streets, without prior agreement with relevant administrative bodies, is something I’ve found myself calling “mobility spam”. While not every bike-sharing outfit is dropping in without warning, there are enough to create a public nuisance as bikes are left in rights of way, dumped in bodies of water, idly propped against a tree, building or mailbox, or inhabit storage racks intended for private use — burdening common public amenities with another proprietary transport system. In the words of one Amsterdam city councillor upset at these schemes’ impact, “Public space belongs to everyone.”

Beyond taking up space, they butt up against other norms as well. Systems in Seattle and elsewhere in the US have run afoul of helmet laws, for instance. In other places, the deploying companies, like their ride hailing cousins, have encountered systems sanctioned or paid for by City Hall. Not a few Dutch cyclists have wondered aloud why a country already awash with 13 million bikes (which is almost one per person, in a population of 16 million) as well as hundreds of rental sources, a reasonably well functioning bike rental extension of the NS (national rail) and where a rich set of tacit rules around cycling have existed for decades, would need a sudden influx of new bikes. As an immigrant from the US, first to the UK, and more recently to the Netherlands, I can tell you, national mobility cultures are real and tangible. Each has its own mobility subcultures which deserve recognition and some respect.

One more bike at Amsterdam’s Muiderpoort Station, one of the city’s busier commuting hubs. Image: author

Credit, or coercion?

The operating models for some of these bike schemes are, naturally, based on a combination of apps, codes and GPS, used for rental, payment, tracking and fleet management. This is similar to the apps used to manage various drop-anywhere car sharing programs. The companies that operate these systems start with a model of how they think customers will behave, and refine pricing, incentives and fleet logistics over time based on learning from usage. Assumptions are built in around customers’ economic behavior and tweaked as the system grows (or falters).

Some, like oBike and Chinese scheme Mobike, the latter available in Manchester since late June, go a step further and include a social rating scheme as a means of nudging users toward system-optimal behavior. The points-based system goes beyond peer-rating made familiar by Uber, Lyft, Airbnb and so on. A customer’s behavior is rewarded or penalized. Like a bank credit rating, lower points mean higher fees.

This social credit system also rewards customers for reporting bad behavior by other customers, adding points for logging another bike improperly parked or otherwise ill-used. “If you see any illegal or poorly parked Mobike, please send us feedback and you will be rewarded with Mobike Credits,” says the company FAQ. A key feature of this is what’s called the Demerit System, which collates reports of bad behavior from fellow riders as well as unspecified “government control agencies”. Leave a bike unlocked while being “intercepted by the police,” and points are deducted, using Mobike’s example.

There is more than a tinge of the control culture of Mobike’s home market in the way its credit system is expressed. It’s no coincidence that China is the home of other newly emergent social credit systems, from ANT Financial’s Sesame Credit to various other schemes being trialled or planned across the country. These systems commonly blend some level of financial scoring with rewards for compliance with civil and social norms preferred by government at different levels. As these systems grow, privacy and transparency advocates express growing concern with the widespread impact of negative ratings. These function as something like an ASBO, with financial, as well as civic and social, implications.

Bigger metal, bigger consequences

Imagine now we aren’t talking about bike sharing, but something a bit larger, heavier and more consequential, like cars. Rapid, and very competitive, development of machine learning and AI has been a big story for the Chinese economy. Players like Baidu are grabbing headlines for their autonomous vehicle projects, with Baidu in particular announcing it will open source its platform. As with AI and associated algorithms developed anywhere, these platforms will have some level of underlying cultural assumptions and norms baked into rule sets. In particular, systems that rely on smooth, safe interaction with other independent actors, as mass mobility does, are designed in part around what is normal, legal driving behavior, as well as the range of possible human behaviors, weather, design of infrastructure, etc.

First of many urban delivery bots. Image: Paul Wasneski / Flickr (public domain)

Alternatively, imagine a boom in drone-based or curb-crawling cargo platforms promised by various pundits and entrepreneurs. One of the drivers that has fueled the explosion of dockless bikes is the ability to source high volumes of bikes cheaply from Chinese manufacturers (also resulting in large piles of the same bikes, rapidly decommissioned), and the availability of low-cost electronics to support the app add-ons. This combination of what author and critic Will Wiles called the “Alibaba-nomics” of pairing “injection moulding and inexpensive motors” is one candidate for the next Detroit, but for hundreds of form factors of mobility — cheap to make, distribute, deploy, and walk away from if the investment fails. Think hoverboards, but without a closet to be dumped in. Given the so-called permissionless model of introducing new software-based models of assumed human behavior presaged by the Internet of Things, this heterogeneous mobility spam future feels more likely, at least in the early stages, than a cleanly organized, standardized ecosystem of well behaved vehicles.

(As an aside, also consider the number of flash-in-the-pan mobility schemes that will most likely take a credit card number, background app access and/or a fragmentary history of travel journeys to whatever startup smelter they end up in.)

This reflection isn’t specific to Chinese products or services alone — here they stand in for fast-moving, “smart” bits of infrastructure that carry encoded cultural values. As systems progress, they will be charged with making legal, ethical and moral decisions while taking on more autonomy — a challenge that is summarized in the oft-cited “Trolley Problem”. Think about how, beyond bikes, autonomous vehicle operating systems will encompass not only mechanically or legally optimal understanding, but economic and social as well. What constitutes efficient driving? Economical? Courteous? Compliant? And for which culture will these be relevant — that of the technology’s design, operation, ownership, presence, or something else entirely? And how might these agree or conflict with the models managing the vehicle approaching at speed on the motorway or side street?

The problems created by the rapid roll-out of proprietary bike-sharing in crowded cities is probably short-term and solvable through a negotiation between existing legal frameworks and the economic interests of the vendors (To be clear, I’m *strongly in favor* of sensible approaches that will improve transport options for the many, operate sustainably, and respect the commons). The queasiness of clumsy social credit systems feel like a layer of poorly implemented gamification. Both are probably issues to be wrangled in the short term. Both, however, are signpost issues that will only grow in complexity and consequence as more muscular mobility technologies are deployed. As Uber found out with its own self-driving car trials, a full-sized SUV can break a lot when it moves fast.

There will be negotiation and smoothing of many operational differences between systems, as there was a century ago when cars from many manufacturers hit the road. But the intelligence that operates these vehicles is being asked to do more, so it must make more decisions and apply rules for its passengers and others around them, decisions and rulesets we won’t always see. Today’s mobility spam sets the stage for transport malware tomorrow if we don’t take a broader view of its various sub-genres, trajectories and impacts. We can ill afford dead metal, singed plastic and junked code in such important social, economic and technical systems. This would be a good time to pay close attention to the early signals so we can anticipate the obstacles and risks that may emerge later and make adjustments today.


Thanks to Madeline Ashby, Susan Cox-Smith, Sjef Van Gaalen and Stephan Hügel for their eyeballs and ideas.