If you walk through any major city today, you’ll encounter one of two realities of micromobility systems: a row of bikes or e-scooters parked in stations, or a mix of free-floating dockless vehicles scattered across the pavement.
This juxtaposition reflects a longstanding city dilemma: upfront investment in station-based systems that offer rider predictability and tidy organisation and free-floating fleets that offer the promise of broader, more flexible coverage but can clutter sidewalks and require costly redistribution.
“Bike share started as a purely station-based model and for good reason,” Caroline Samponaro, Head of External Affairs at shared micromobility operator Lyft Urban Solutions (LUS), tells Zag Daily.
“Stations gave cities predictability, transit integration and operational control. The introduction of free-floating came with the promise of lower costs and less public space consumption but what we’ve learned since is more nuanced. We see that free-floating programmes consistently underperform station-based ones on utilisation, create compliance and safety burdens and generate hidden costs cities eventually have to absorb.”
Going hybrid
This is where a quiet parking revolution is changing the landscape.
A new study conducted by Toronto’s York University in conjunction with LUS revealed that the future of urban transit isn’t an ‘either/or’ choice between docked or dockless – it’s a carefully optimised hybrid.
According to Samponaro, the market is moving toward station-based systems with hybrid capability where flexible parking is a “complement to the station network rather than a replacement for it.”
Under the hybrid system, a commuter can unlock a bike from a physical station located next to a transit stop and park it in a designated out-of-station parking zone near their home.
Hybrid networks that combine the benefits of physical stations and out-of-station parking zones enable four distinct trip categories that traditional networks built around a single parking type cannot support: station-to-station, zone-to-zone, station-to-zone and zone-to-station.
Cities across the world are already proving that these systems thrive where others struggle. In San Francisco, the Lyft Bike system integrates a traditional physical station network with designated parking zones for e-bikes, while Madrid leverages stations in core areas while providing riders the option to park e-bikes outside of stations in the system’s periphery.
Testing infrastructure scenarios at scale
The study from York University and LUS utilised a mix of publicly accessible open-source data, artificial intelligence, machine learning and location data to simulate four distinct 2026 expansion scenarios for bike share in the city of Hamburg, Germany.
Samponaro notes that Hamburg was selected because it is “a large, expansive city where a hybrid model combining station-based infrastructure and out-of-station parking options is the most logical approach.”
The study aimed to address Hamburg’s current blank-spot districts and bring bike share to these underserved areas while maximising utilisation and ensuring robust links to the public transit network.
The existing system (Scenario A) was pitted against three expansion models, each utilising a similar scale of new vehicles but configured with different infrastructure layouts:
- Scenario B (Dock-only expansion): Added 20 transit-oriented docking stations across high-activity corridors.
- Scenario C (Hybrid, preferred): Added 10 docking stations at high-density hubs and 10 out-of-station parking zones in the surrounding periphery.
- Scenario D (Reverse hybrid): Placed 10 out-of-station parking zones in the dense city core and relegated 10 docking stations to the low-density periphery.
The results were telling. Both the Scenario B (dock-only) and Scenario C (hybrid, preferred) models successfully boosted annual ridership by five to six percent relative to Scenario A, pushing ridership from 1.22 million trips to 1.30 million and 1.29 million trips, respectively.
However, the capital expenditure (CapEx) required to achieve these gains should be taken into account. By leveraging out-of-station parking zones for half of the net-new network, the hybrid Scenario C achieved 99% of the ridership benefits of Scenario B with only half of the upfront capital cost.
“A hybrid model combining station-based and out-of-station parking can deliver the best results in terms of system performance,” says Samponaro.
“But the critical nuance is that placement matters as much as model type: station-based infrastructure belongs in high-demand, transit-connected locations; out-of-station parking zones work better in lower-density areas.”
Scenario D (reverse hybrid) was not recommended. After the docking infrastructure was removed from the dense city centre and relocated to the outskirts, the estimated ridership fell to 970,000 annual trips (20.2% of the original baseline). Despite this significant decrease in utility, Scenario D still incurred a higher implementation cost (+110% CapEx) compared to the hybrid model.
The simulation was then extended to five other global markets – Paris, New York City, London, Copenhagen and Montreal – where the hybrid model consistently captured the greatest share of potential ridership.
Samponaro emphasises that “removing core station-based infrastructure and replacing it with purely out-of-stations parking areas leads to a dramatic decline in ridership, highlighting the importance of stations in dense urban cores.”
While giving riders a flexible way to end their trip can complement station-based systems, a purely dockless system is not an effective substitute where demand is highest. This is backed up by the behaviour of riders in the real world.
“Across LUS-operated systems with hybrid parking, nearly 70% of rides still start and end at a dock and 88% involve a station at one or both ends,” Samponaro reveals.
“Riders overwhelmingly prefer to start a ride from a station and that preference is what drives ridership volume and keeps programmes financially sustainable.”
BikeShare Pro: The brain of the city
Historically, city transportation agencies have expanded their bike share networks through ad hoc planning based on observed demand, available funding or incremental station additions.
But modern systems require a more measured approach that jointly considers ridership potential, accessibility, transit connectivity, equity objectives, operational performance and lifecycle costs.
This is why Dr. Mehdi Nourinejad, Assistant Professor in the Department of Civil Engineering at York University, co-developed the BikeShare Pro planning and optimisation platform with Dr. Ghazaleh Mohseni in 2024.
The white paper is the result of a research partnership between LUS’s deployment across 55+ systems and Dr Nourinejad’s analytical team.
“BikeShare Pro bridges the gap between long-range micromobility planning and the day-to-day tactical decisions cities and operators must make as bike share systems evolve,” says Dr Nourinejad.
“Many cities have access to large amounts of trip, station, demographic, transit and operational data but they often lack an easy-to-use platform that turns those datasets into fast, actionable planning insights.”
By breaking a metropolitan area down into standardised hexagonal grids, the platform allows planners to run predictive scenarios. Users can also adjust a “policy weight” slider to automatically generate optimised layouts based on their immediate goals, whether that’s concentrating infrastructure in high-demand areas to maximise raw ridership or prioritising equity, suburban accessibility and public transit integration in a city’s outskirts.
These planning insights don’t stop at the initial network design. As cities evolve, from expanding coverage to shifting toward electric fleets, the same data-driven approach applies to what comes next.
Designing for large electric fleets
As cities transition to electric fleets, the argument for keeping stations as the network’s foundation becomes even stronger.
“Electrification will accelerate the shift toward station-based infrastructure,” Samponaro forecasts. “As e-bike fleets grow, in-dock charging becomes less of a differentiator and more of an operational necessity. Every bike trip that ends out of station results in a bike that may need a manual battery swap. When a charging station exists in the most popular parking locations, the shared e-bike fleet charges while not in use, improving e-bike uptime and reducing operational cost and complexity.”
This is one of several factors pushing cities toward a single-operator model. Big cities like Madrid, Barcelona, and Brussels have already moved to restrict or eliminate independent free-floating operators, instead consolidating around public, docked systems. A desire for order on the pavement and more efficient use of capital investment are cited as clear rationales for this shift.
“Cities are finding that experimenting with multiple operators sharing the same geography produces worse outcomes,” says Samponaro. “It leads to disorganised pavements, lower utilisation, higher operational costs and a less appealing brand for potential sponsors, with cities left managing clutter and enforcement of parking rules instead of expanding networks,” says Samponaro.
“A single public system with a well-designed hybrid model is emerging as the dominant framework. This maintains the objective of a public bike share programme that ties into the city’s infrastructure and planning while allowing for the benefits of innovations in tech that have surfaced in recent years to benefit the public approach.”
As cities strive to reach ambitious low-carbon transportation goals, the white paper demonstrates that the most resilient and cost-effective networks in the future won’t rely on excessive infrastructure spending or unlimited fleet growth.
Instead, they will depend on data-driven planning – showing that bike share’s future isn’t a trade-off between reliability and flexibility but about developing a robust station network capable of adapting to a city’s needs.
