EDITOR’S NOTE: The following is an excerpt from “The Urban Improvise: Improvisation-Based Design for Hybrid Cities,” by Kristian Kloeckl, published by Yale University Press. In it, the author unveils an urban design theory grounded in digital connectivity and based on unscripted behavior that in many ways resembles improvisational dramatic performance. Although of late, dockless bike-share companies have over-expanded in some markets and then abruptly closed shop, and issues have persisted about mechanical problems and the unregulated nature of the pickup and drop-off system, Kloeckl argues in this excerpt how the impromptu essence of park-where-you-will dockless bike-share is actually a positive expression of responsive urban design.

Today’s latest generation of bike-sharing systems have done away with the docks, consisting in the bikes alone. Bikes can be locked whenever and wherever the user completes a trip. The bikes are physically scattered throughout the urban territory without any planning, and yet they remain connected to a remote sensing platform. This latest breed of dockless bike-sharing systems builds on a remarkable history of bike-sharing modalities that goes back half a century.

In 1965, the Dutch counterculture movement Provo proposed the White Bicycle Plan to solve Amsterdam’s inner-city traffic problem. The plan was for the city of Amsterdam to purchase twenty thousand white bikes annually to be distributed unlocked throughout the city. They were to be considered public property and free for everyone to use at will. When the city of Amsterdam did not take up this plan, Provo went ahead alone, painted fifty bikes white, and distributed them around Amsterdam’s inner city, unlocked and free for everyone to use.

The white bikes were soon taken out of circulation by citizens for private use, as well as by the police, who confiscated white bikes in circulation on the basis of legislation that requires bikes to be locked when parked in public spaces. The bikes were returned to the Provos, who, in a second attempt, installed number locks on all remaining bikes, painting the numeric code onto the bikes to bypass the legal impasse. The system did not live on for long, however, as bikes were frequently damaged and stolen, and the White Bicycle Plan soon came to an end .

The first large-scale bike-sharing systems were installed in 1995 in Copenhagen (the ByCyklen project) and in 2001 in Vienna (the Gratisstadtrad project first, followed by the Viennabike project in 2002). Both systems used dock stations, painted bikes, advertisement on the bikes to finance operation, and a coin-based docking mechanism similar to some supermarket shopping carts still in use today. A coin was used to unlock the bike at the dock, and when locking the bike at another docking station, a coin of equal value was retrieved.

Vélib in Paris launched the third generation of bike-sharing system in 2007, operated as a concession by the advertising and street furniture company JCDecaux. The innovation that set this third generation of bike-sharing systems apart was the requirement of a personal identification and a higher deposit for the use of the bikes. In order to register with the system and unlock the bikes, the use of a credit card or a cellphone number was required. Holding a larger deposit through the credit card and requiring personal identification helped these systems avoid vandalism and theft of the bikes.

Today’s dock-based bike-sharing systems operate essentially in the same way as the Vélib system. The main components are docking stations with wired or wireless telecommunication, a kiosk interface to use with a bank card for registration with the system, and bikes that contain a docking mechanism that is mechanical and contains a digital connection to identify the specific bike connected at each dock.

A key issue for the operation of these systems is the so-called balancing of bikes between the dock stations throughout the day. Trucks move bikes from stations with excess numbers of bikes to other docks with too few bikes. The purpose of this is to balance the system and counteract dynamics by which many people use the bikes to ride from outer districts to inner-city locations in the morning and then the other direction during the evening commute. The precise strategy for rebalancing is complex and also considers predicted demand patterns based on the analysis of historical data.

The latest generation of bike-sharing systems from the past couple of years has introduced a significant innovation to the bike-share operation. It brings bike-sharing surprisingly close to the operational practice of the first system of White Bikes in Amsterdam. This new generation of bike-share system does away with the use of docking stations throughout the city; the system is dockless. The first dockless bike-sharing operator, ofo, launched in 2015 in Beijing after the founding of the company in the previous year by members of the cycling club of Peking University. The company has since expanded to numerous cities internationally. Also, competing dockless systems have entered the scene, such as the companies oBike, LimeBike, and Reddy.

These dockless bike-sharing systems, as the name implies, operate without the need of physical docks to retrieve and return the bikes, as is the case of dock-based systems. The dockless system consists only in the bikes themselves. Bikes are parked anywhere in the city, and trips can start and end at any location. Unlike Amsterdam’s early White Bikes from 1965, however, the bikes are now tracked by a sophisticated use of multiple systems. The bike location is tracked by a GPS system, and the lock mechanism is activated through the operator’s smartphone app. In use, pointing the smartphone’s camera at the bike lock’s graphic QR code is sufficient to unlock the bike and pedal away. In order to conclude the ride, parking the bike in any public location and locking the mechanical lock is all that is required. Finally, using the operator’s online map in the smartphone app helps to locate parked available bikes that are not in direct sight from a user’s location.

The way these kinds of systems work technically varies slightly from system to system and from town to town. Some bike locks contain built-in GPS and cellular units. These systems always know the location of all bikes at any moment, during a ride and when parked. Other systems only have a Bluetooth connection in the bike lock and make clever use of the user’s smartphone to get GPS location and cellular data transmission via the smartphone’s Bluetooth connection. In this case, the system only receives the location during the locking and unlocking of the bike.

Where to park the bike is entirely up to the users in these systems, and this clearly provides room for interpretation and conflict. Several operators, such as oBike and LimeBike, give credit points to users for returning bikes to designated bike parking locations. The credit system deducts points for parking the bike in illegitimate locations and for not locking the bike. Credit is also given for reporting bikes that are illegally parked or broken.

A System Designed for Openness

The dockless bike-sharing system appears to be characterized by a high degree of openness and invites users to take initiative. During everyday operation, and once you are registered with the system, whenever you see a yellow bike (in the case of ofo bikes) parked, you can walk up to it, unlock it, and pedal away at will.

The system reduces the cognitive and operational overhead. Riding a bike comes close to the immediacy of Amsterdam’s White Bikes. When you see a bike in plain sight or virtually on a digital map, you walk up to it, unlock it with your smartphone, and start riding it, to then leave it parked at any spot. It is a strength of the system that it is so inviting and easily accessible; people come up with new ways of integrating these bikes into their daily routines that they themselves did not consider before.

The openness of the dockless bike-sharing system lies in that it offers enormous room for interpretation of how to appropriate and integrate the service within one’s everyday routine: ad hoc acceleration of trips that add the convenience of a bike to what was planned to be a walk or public transport trip or the use of a bike alone or together with friends for a bike tour of an entire day, without the need to return to any specific location, are just some of many possibilities. While writing this book, I saw a little boy ride his own bike, accompanied by his parents both on yellow ofo bikes. A family bike ride was made significantly more accessible because of the availability of these bikes. The bikes are painted in bright colors so that they stand out visually. Everybody who rides such a bike becomes an apparent model for others to see with regard to what can be done with the bikes, where they can be parked, and where they can be found.

The structural vitality of the project comes from the effective orchestration of functionalities from networked technologies and system governance. The GPS of the user’s smartphone locates the bike during the unlock operation, and the phone also provides connectivity to communicate the bike’s location and the user’s identification to the bike operator’s system. The QR code, scanned by the phone’s camera, has finally found a worthwhile use, after having been around for years in search of a meaningful application. Implemented in this way, the QR-based bike-unlock mechanism turns a limit of the QR code into a feature: the requirement of a direct line of sight to visually scan the code ensures that whoever unlocks the bike is physically present at the location of the bike; a remote unlocking by users becomes impossible due to the deliberate use of the constraints of a technology.

The fields of possibility in this bike-share system are quite strikingly circumscribed by the omission of elements. Taking the docks out of the equation, the field of possibility is continuously re-created by the users of the system and their choices of where they park the bikes. This is a significant change to pay close attention to: the users of the system cause a continuous redesign of the system itself. This is not a system based on balancing average ridership demand, as is done in the case of dock-based systems.

Here, every single decision of where to park a bike matters, and each decision impacts the system significantly with regard to what possibilities are offered. A bike might just be parked literally outside your door one day and offer you a ride.

There are limitations to the openness of the system, however. Dockless bike-sharing systems largely require the possession of a smartphone as well as a credit card for payment. Some exceptions exist. LimeBike in Seattle, for example, introduced cash payments in 2017. People can prepay in cash at a physical office in the city and then call a number from any phone to have a bike unlocked. These systems also tend to be limited to municipal boundaries; they rely on agreements with municipalities. In 2018, thousands of dockless bikes were being deployed in towns around Boston. “Notably absent from that list [of towns] are Boston, Brookline, Cambridge, and Somerville—the four communities that use the Hubway system. Their agreement with the private company that operates Hubway bars other bike-share systems,” comments the Boston Globe. This is an interesting new territorial condition that is created by commercial agreements and that is defeated by the very nature of bikes being mobile enough to cross municipal boundaries anytime.

Rebalancing the system becomes even more of a challenge with dockless bikes than with their dock-based equivalents, as the individual bikes are spread out and often hard to locate. The system does get unbalanced, with bikes being rare near subway stations during the evening commute, as users are keen to use them for the last mile of their commute. Rebalancing strategies, however, do exist and are either truck based or user based via dynamic pricing/crediting schemes.

A System Both Timely and Serendipitous

The quickness of the unlocking operation in bike-sharing systems matters. Dockless systems are very quick. You point your phone’s camera at the lock’s QR code, and in only a couple of seconds the lock opens. The immediacy is essentially at the level of Amsterdam’s White Bikes from 1965—we have come full circle. There is no looking for the nearest bike dock, no payment or other phone interaction needed, and no kiosk interaction. The quickness of the interaction and the physicality of it matters in this service. The system uses sophisticated digital networks, but it also uses the physicality of the bike and its lock as an effective interface.

There is no need for a user to decide upfront on the destination of the trip. There is no need to identify dock locations at the end of the ride and whether dock spots are available to park the bike once arrived. The timing of the system is based on kairos, the opportune moment. You see a bike and decide there and then to use it. There is nothing else to consider to make this work operationally; it is an effective low-stakes commitment.

Rhythms, ever-changing locations of the bikes, emerge where bikes can be found at different times of the day. As a user of the system, one becomes acutely aware of the rhythms of fellow bikers using the system and the flow of bikes from one part of the city to another and back again.

When parking the bike, a user becomes aware that the place the bike is parked will condition the next use of the bike: whether it will be found easily and whether it is parked at a place with much foot traffic. It is an element of anticipatory design that invests and involves the user directly. Essentially, as a user, you design the future findability of the bike and the point of access to the system on every ride.

A System That’s Visible and Always Changing

The dockless bike-sharing system is in constant flux. Bikes are parked in different locations all the time, and the system, in this way, is capable of adapting to changing conditions of traffic, housing, residency, and so on. Precisely because there are no predetermined docking stations, the parked bike locations remain relevant in the context of constantly changing urban conditions, as the locations are an expression of actual use.

Cues about the state and the behavior of the system are provided directly by the visibility of the bikes in the urban environment. Bikes are very visible, as in the case of the bright-yellow ofo bikes. It is easy to spot them when they are nearby, and it is easy to spot accumulations of several bikes at specific locations at certain times (such as subway stations, schools, and office buildings). It is easy to observe also that at these same locations, there can be no bikes at all at other times.

The system is understood by the physical and visual presence of the bikes in urban space when parked as well as when ridden. You become aware of these bikes visually when they ride by you. You see these bikes when they are parked on the sidewalk or at other public spots. You use the system when you ride the bike, and you observe the system when others do so. When observing the system, you are also using it, as you learn from observing the ever-changing locations and concentrations of the bikes in your immediate environment.

Parking dockless bikes is fundamentally different from parking dock-based bikes. Where to park becomes an open question for the biker. There are indications of where to park: rules from the operator that urge users not to block the pedestrian way of passage. In the real context of sidewalks, finding an adequate spot to park leads users to become aware of the intricacies of public use of any particular spot. Parking these bikes is an excellent example of situated action, as you find yourself negotiating the affordances of tight sidewalks, lamp posts, and other elements to park the bike, abiding by the operator’s rules as well as negotiating the contingencies of the actual characteristics of place.

Accountability beyond the bikes in plain sight is provided through the tracking system and the online maps. The bikes are GPS located, and wherever they are parked, they can be located. The trips of the bikes are also linked to the identity of the user, and thus there is accountability for potential damage to the bike, parking in spots deemed inappropriate by the operator, and similar criteria.

A System That Expects the Unexpected

Today’s dockless bike-sharing systems are wild systems. They are similar to Amsterdam’s White Bikes. It is clear that anything can happen to these bikes, and a lot already has happened to them, from being “parked” on top of trees to being recovered from rail tracks and rivers.

Bikes that do not have a built-in GPS unit can be transported to a different location while locked, and in this way, they can get off the grid quite literally. Their actual location no longer corresponds to the virtual one that is maintained in the operator’s system and that is visualized on the digital maps of the system. Cases exist in which bikes have been hidden behind dumpsters and bushes to “reserve” them for personal use at a later point, with users exploiting leeway in the system. Bikes are also frequently driven beyond municipal boundaries and dropped off in neighboring cities that have exclusive contracts with competing bike-share providers. This triggers operators having to recover bikes directly.

These uses are idiotic, [to use the phrase] as described by philosophers Isabelle Stengers and Gilles Deleuze, in that they are acts by citizens that do not play by the publicly accepted rules. They do not contribute to the common cause of the system in the way that the citizen is expected to contribute but instead represent personal interpretations of the system. They “slow things down.” They are, however, by being idiotic, also a valuable manifestation of what is possible.

The otherness of the system is that you never know where to find a bike in advance. You can never be certain to find a bike nearby. That is different from dock-based systems, which are more predictable due to their limited number of bike locations, which concentrates bike availability, and across which operators balance availability. In a very concrete way, you cannot count on dockless bike-sharing systems, and yet they appear to work and fulfill a role. The balancing of the system does not refer to docks but literally to the entire urban territory. Some days a bike is outside your door; then again, on other days, the nearest bike is a several minutes’ walk away. It is a system that does not allow you to settle in a “usual” mode of operation; it keeps surprising you, keeps you on your toes. The operation of dockless bike-sharing systems is fundamentally based on otherness. The system is always other than expected. It is a system that, in its current form, exists beyond control.

Adapted from “The Urban Improvise: Improvisation-Based Design for Hybrid Cities,” by Kristian Kloeckl, published by Yale University Press. Copyright © 2020 Yale University Press. Reprinted by permission of Yale University Press.