A research project of the Lucerne University of Applied Sciences and Arts (HSLU) and the University of Basel, funded by the Swiss Federal Office of Energy

The potential of e-bike sharing for sustainable mobility

Two e-bike sharing systems operate side by side in Basel: the free-floating Pick-e-Bike with S-pedelecs and the dock-based PubliBike Velospot with pedelecs. We linked about 1.4 million rentals with a survey of 1,844 registered users to find out who uses these services intensively, and why.

Duration
December 2023 to January 2027
Funding
Swiss Federal Office of Energy (SFOE), grant SI/502720-01
Institutions
HSLU Institute of Tourism and Mobility; University of Basel, Faculty of Business and Economics
Data partners
Pick-e-Bike and PubliBike Velospot
00:00
A typical day in Basel. Rentals from 26 August to 1 September 2024, routed along the street network and placed on an average day. About 1,600 of the 1,800 routed rentals pass through this view. [1]Pick-e-BikePubliBike Velospot

Published results

Key findings

Practical benefit explains intensive use better than environmental attitudes. Results of the first article, based on 1,743 survey respondents linked to their rental histories.

  1. A small group of users generates most trips

    A Gaussian mixture model on rentals per month divides users into three segments: low (48%, 0.20 rentals a month), medium (39%, 1.60) and high intensity (13%, 12.71). The top 20% of users account for 80.3% of trips (Gini 0.77). Use is more concentrated in the dock-based system (Gini 0.83 vs 0.75), where high-intensity users are 2.4 times as common (23.5% vs 9.6%).

  2. Environmental identity does not lead to more use

    Users with a stronger pro-environmental self-identity are less likely to be in a higher-intensity segment (OR 0.82, p = 0.01). One plausible reason: in Basel, people with strong environmental attitudes already walk, cycle or take the tram, so a shared e-bike adds little for them.

  3. Combining with public transport is the strongest driver

    Using e-bike sharing together with public transport or walking, or as a fallback, raises the odds of higher-intensity use the most (OR 1.53, p < 0.001). Doubts about availability and cost are the strongest brake (OR 0.61, p < 0.001).

  4. The two systems have different frequent riders

    In the dock-based system, frequent riders are mostly highly educated and live in the urban core. They ride fixed commuting routes on which the bike is faster than the tram (perceived advantage over public transport: OR 1.53). In the free-floating system, frequent riders are younger than occasional users and less often own a car (48% vs 69%). They combine the service with other modes (OR 1.74). In the article we call them routine optimizers and flexibility maximizers.

  5. Shared e-bikes mostly replace public transport and walking

    75.5% of users at least sometimes replace a public transport trip, 50.9% a walk. Of users with access to a car, 33.9% sometimes replace a car trip (free-floating 36.4%, dock-based 22.0%). E-bike sharing mainly relieves and extends public transport. It is not primarily a car substitute.

Open question: does sharing lead to buying an e-bike?

Owning a private e-bike goes with less shared use (OR 0.73, p < 0.01). About half of the 629 owners and prospective buyers say that e-bike sharing contributed to their decision, more often among high-intensity users (63% vs 46%). With cross-sectional data we cannot tell whether sharing causes the purchase.

Low: 0.20 rentals/month Medium: 1.60 High: 12.71
Usage segments. Share of the 1,743 linked users per segment and mean rentals per month (Gaussian mixture model).
Which modes shared e-bikes replace. Flow width is the number of users who at least sometimes replace public transport, walking or the car (free-floating n = 1,382, dock-based n = 462; Table 3 in the article). *Car: share among users with access to a car.

What predicts usage intensity

Odds ratios from the ordered logit model (N = 1,743, McFadden pseudo-R² = 0.104). Values above 1 raise the odds of belonging to a higher-intensity segment, values below 1 lower them.

Attitudes and perceptions Mobility tools Demographics Place of residence
Points show odds ratios, bars 95% confidence intervals; hollow points are not significant. All 15 predictors of the model are shown, as published in Table 8 of the article. Continuous predictors are standardized.

Reading the model

Combining e-bike sharing with other modes is the strongest positive predictor, concerns about reliability and cost the strongest negative one. Environmental self-identity is negative: a stronger green identity goes with about 18% lower odds of a higher-intensity segment.

A Brant test rejects proportional odds overall, mainly because of control variables (half-fare travelcard, place of residence, income). Among the attitudinal factors only environmental self-identity deviates: its negative association is concentrated at the threshold to high intensity (OR 0.68 vs 0.90 at the medium threshold). A multinomial logit in the appendix of the article gives the same ranking of predictors.

Separate models per system, odds ratios for three attitudinal factors
FactorFree-floatingDock-based
Combination with other modes (F5)1.74 ***1.13 n.s.
Advantage over public transport (F4)1.08 n.s.1.53 **
Environmental self-identity (F7)0.870.77

*** p < 0.001, ** p < 0.01, n.s. not significant. Environmental self-identity is below 1 in both systems.

Implications for practice

E-bike sharing is a useful but limited part of the transport system. The results suggest investing in what keeps frequent riders riding.

Integrate with public transport

This is the strongest lever: combined fares or subscriptions, availability shown in journey planners such as the SBB app, and stations or parking zones at transfer points.

Make availability reliable

Doubts about availability and cost hold frequent riders back more than anything else. Fleet size, rebalancing, real-time information and short reservations (for example 15 minutes) address exactly these doubts.

Promote practical benefits

Frequent riders are not the most environmentally minded users. Time savings and good connections to public transport are the stronger arguments.

Measure the right benefits

Most replaced trips were already made by public transport or on foot, so CO₂ savings are a weak yardstick. Relief for crowded public transport at peak times, last-mile connections and the fallback role during disruptions matter more.

Green shared bikes parked next to a bus shelter in Singapore
Singapore: shared bikes parked directly at a bus stop. [2]
A worker next to a small vehicle stacked high with shared bikes on a street in Beijing
Beijing: bikes collected for redistribution. [3]
Sharing Zone sign with operator logos at Basel SBB, with Velospot and Pick-e-Bike bikes
Basel SBB: one marked zone for all providers at the station. [4]

Fare integration is now being tested in the Basel region. In the U-Two field test, FHNW, HSLU, BVB, BLT and TNW integrate shared bikes and e-bikes into the U-Abo and the GA travelcard. The test runs through 2026.

Comparing the two systems

The second article compares both systems over their full operating history (May 2018 to March 2025, 1.47 million trips, no survey data). Docking, vehicle type, speed, service area and fares all differ at once, so each system is treated as a package and no single design feature explains a difference.

White Pick-e-Bike S-pedelecs and red PubliBike Velospot pedelecs parked next to a red Sharing Zone sign in front of the Basel SBB station building
Both systems side by side: Pick-e-Bike S-pedelecs (white) and PubliBike Velospot pedelecs (red) in the sharing zone at Basel SBB station, September 2024. [5]

Usage

The free-floating system has a broad, mostly occasional user base: 26,744 active users, median trip 2.05 km, Gini 0.72. The dock-based system has fewer but more frequent users: 6,519 users, median trip 1.44 km, Gini 0.80. 93% of its activity is in Basel-Stadt, and its users are younger and more gender-balanced.

Space

Free-floating trips follow where people live and thin out with distance from the centre. Dock-based trips follow where people work. With time window and area matched, the ratio of trip distances falls from 1.49 to 1.19, so the gap is partly a matter of coverage.

Weather

Rain reduces the free-floating morning peak the most (IRR 0.67 per mm), while temperature hardly matters there. Dock-based demand follows temperature throughout the day.

Scale

Together, both systems amount to about 0.24% of BVB boardings. Free-floating trips increasingly start in areas with medium public transport quality (classes B and C, from 34% to 42%). Dock-based trips stay concentrated around the main hubs (class A, about 80%).

Four panels: heatmaps of trips by hour and weekday for the free-floating and the dock-based system, and bar charts of the monthly share of annual rentals for both systems
When the systems are used. Share of trips by hour and weekday (A, B) and by month (C, D). FFEBSS: free-floating Pick-e-Bike; DBEBSS: dock-based PubliBike Velospot. [6]
System indicators, 2024
IndicatorFree-floatingDock-based
Annual rentals203,738107,156
Active users12,6583,778
Trips per active user16.128.4
Mean trip distance (km)2.621.57
Mean trip duration (min)17.311.4
Gini, trips per user0.720.80
Median user age (years)39.133.0
Trips per active vehicle-day2.301.84
Median fleet idle time (min)384613

What this means for planning

The two systems serve different purposes and should not be judged by one benchmark. A single benchmark undervalues either the regional reach of the free-floating system or the intensive use of the dock-based one. Their small volume relative to public transport makes large-scale substitution unlikely. Whether they complement public transport cannot be answered from operational data alone, as these contain no trip purposes.

Source: Stiebe, von Arx and Sager-Müller (2026), Journal of Urban Mobility 10, 100288, open access. Based on all operational records, so the figures are not directly comparable with the survey sample of the first article.

A field experiment: speed setting

In August 2024 Pick-e-Bike began offering its S-pedelecs in a throttled mode, as a pilot project with the Federal Roads Office (ASTRA). POTEBS follows this intervention in the operational data. It is a rare field experiment: the same fleet in the same city, with the licence barrier removed for part of the user base.

Rear view of a Pick-e-Bike S-pedelec with a round rear light showing the number 25, above the yellow licence plate
The throttled mode is visible from behind: a rear light showing 25 marks a bike that is limited to 25 km/h. [7]

What changed

Since 20 August 2024 the operator can limit an S-pedelec to 25 km/h of pedal assistance and 400 W by software. At that speed the vehicle counts as a bicycle under Swiss law, so riders no longer need a driving licence and can register with an identity card from the age of 16. A separate rear light marks the slow mode, and the operator requires a helmet on all its vehicles. Riders with a licence can still use the 45 km/h mode. Details are on the operator's speed-setting page and on the project page.

Why the licence matters

The licence requirement excludes a large part of the population. In the urban core of the service area 30.1% of adults hold no car driving licence, in the wider service area 25.5%. Among 18 to 24 year olds it is 62.1% in the core and 53.8% in the catchment area (Mobility and Transport Microcensus 2021).

More use per bike

Rental minutes per active bike per day stood at about 12 to 13 before the launch and reached 33 in October 2025, around 2.5 times the pre-launch average. The effect built up over months rather than appearing at once, which speaks against a pure novelty effect.

A younger and more regional user base

The median age at registration fell from 33.6 years before the launch to 30.8 years after it, and registrations rose. The share of new customers living outside Basel-Stadt grew from 63% to 67%. Speed-setting users stay active for a slightly shorter period than regular users (median 50 versus 57 days).

The fleet becomes the limit

Local shortages increased. The share of active hours in which demand exceeded the bikes present, or in which no bike was present at all, rose from about 7% to 12%. The former evening peak broadened into a daily commuting pattern, which concentrates demand in the same hours.

No sign of cannibalisation

In the eight months after the launch, Pick-e-Bike rentals in the city rose by 26% and those of the dock-based system by 86%. Both systems grew, which points to an expanding market rather than displacement. The comparison is descriptive: the dock-based records end in March 2025 and that system was expanding its network at the same time.

Use per bike before and after the launch. Monthly averages from August 2023 to March 2026: rental minutes per active bike per day and total fleet hours per day. The dashed line marks 20 August 2024. Hover for exact values. [8]

The pilot was set up with ASTRA for two years and is limited to the cantons of Basel-Stadt, Basel-Landschaft and Solothurn; ASTRA receives regular progress reports from the operator. That first period ran until August 2026, and ASTRA has allowed the trial to continue. We analyse the trial with the operator's records and report to Pick-e-Bike every quarter. The findings above were presented at Thredbo 19, the international conference on competition and ownership in land passenger transport, in Valletta on 15 September 2026.

Data explorer

Four views of the operational data: service areas, one week of routed trips, where trips start and end, and how unevenly use is distributed across users.

Basel is a demanding test case: it has the highest public transport share in Switzerland (32.5% of trips), a cycling share of 11.7% and the lowest car ownership among the large Swiss cities (319 cars per 1,000 inhabitants). Both systems have operated side by side since September 2021. The map shows the municipalities served by each system and their overlap; PubliBike Velospot stations can be added as a layer. Dashed outlines mark the Pick-e-Bike pilot in Lausen, Itingen, Sissach and Gelterkinden (June to November 2026, one station each).

Publications and presentations

Journal articles

  • Stiebe, M., Krysiak, F. C., von Arx, W., & Weggelaar, B. (2026). Pragmatism, not ideology: drivers of e-bike sharing usage intensity. Transportation Research Part D: Transport and Environment, 161, 105621. doi.org/10.1016/j.trd.2026.105621 (open access, CC BY 4.0)
  • Stiebe, M., von Arx, W., & Sager-Müller, S. (2026). Contrasting usage and spatial profiles of free-floating S-pedelec and dock-based pedelec sharing in Basel: A longitudinal system-level comparison. Journal of Urban Mobility, 10, 100288. doi.org/10.1016/j.urbmob.2026.100288 (open access)

Conference papers and talks

  • Stiebe, M., & von Arx, W. (2026). Market expansion through performance throttling: an analysis of systemic friction, user rejuvenation and the occupancy trap in shared micro-mobility. Thredbo 19, International Conference Series on Competition and Ownership in Land Passenger Transport, Valletta, 15 September 2026.
  • Stiebe, M., & Weggelaar, B. (2026). Drivers of e-bike sharing usage intensity: evidence from free-floating and dock-based systems in Basel. Talk, Sustainable Future Research Lunch, University of Basel, 20 March 2026.
  • Stiebe, M. (2025). Pragmatists, not idealists: what drives e-bike sharing usage intensity and why car substitution remains elusive. Talk, Economics Lunch Seminar, Faculty of Business and Economics (WWZ), University of Basel, 17 December 2025.
  • Stiebe, M., & von Arx, W. (2025). Understanding user behavior in dock-based and free-floating e-bike sharing systems: a multi-method study from the Basel metropolitan area. mobil.TUM 2025, 14th International Scientific Conference on Mobility and Transport, Nanyang Technological University, Singapore, 24 to 26 November 2025.
  • Stiebe, M., & von Arx, W. (2024). Comparative analysis of user characteristics and use patterns in free-floating and station-based e-bike sharing systems: empirical insights from the Basel metropolitan area. European Transport Conference 2024, Antwerp, 17 to 20 September 2024. aetransport.org
  • Stiebe, M., & von Arx, W. (2024). Comparative analysis of user characteristics and use patterns in free-floating and station-based e-bike sharing systems: insights from the Basel metropolitan area. Swiss Transport Research Conference 2024, Ascona, 15 to 17 May 2024. PDF
  • von Arx, W., & Stiebe, M. (2024). Obstacles to economic sustainability of free-floating e-bike sharing systems: a Basel-based case study. Talk, Swiss Transport Research Conference 2024, Ascona, 15 May 2024.

Media

  • «E-Bike-Sharing boomt in der Schweiz!» News article, Lucerne University of Applied Sciences and Arts, 13 November 2024. hslu.ch

Related links

About the project

How can e-bike sharing support the mobility transition as part of a flexible, multimodal transport system?

  1. Which user groups use e-bike sharing, and in which multimodal contexts? What hinders use, and which benefits do users report?
  2. Which business models for e-bike sharing, alone or combined with public transport, could build on these user needs?
  3. How can e-bike sharing contribute to sustainable mobility in urban and peri-urban regions, and which measures support this?

Background

Studies of bike sharing consistently find that 10 to 22% of registered users generate 50 to 81% of trips, and about half never return after signing up. Operating costs often exceed revenues. Whether a system can sustain itself therefore depends on a small group of frequent riders.

Earlier research explains who signs up and what drives total demand, but not which personal characteristics turn a registered user into a frequent rider, or how this depends on the design of the system. POTEBS addresses this by linking about 1.4 million trip records to individual survey responses from two systems operating side by side. Data, survey and models are described on the methods page.

A row of red PubliBike Velospot pedelecs in front of Basel SBB station
Basel, 2024: PubliBike Velospot pedelecs at the station. [9]
Dense rows of yellow and blue dockless shared bikes on a pavement in Shanghai
Shanghai, 2025: dockless bikes in large numbers on the pavement. [10]
A man riding a yellow shared bike across a street in Guangzhou
Guangzhou, 2025: a shared bike in everyday traffic. [11]

Timeline

2023
Dec
Project start
Funded by the SFOE. Trip data from Pick-e-Bike and PubliBike Velospot form the empirical basis.
2024
May
STRC 2024, Ascona
First comparison of user characteristics in the free-floating and the station-based system.
Aug
Speed-setting launch
In a pilot project with ASTRA, Pick-e-Bike limits its S-pedelecs to 25 km/h for users without a driving licence. POTEBS follows the intervention in the operational data.
Sep
ETC 2024, Antwerp
Updated comparison at the European Transport Conference.
2025
Jan
First advisory board meeting
With transport authorities, operators and policy institutions.
Apr
User surveys
Parallel surveys of registered users of both systems: 1,844 valid responses, 1,743 linked to trip histories.
May
Non-user survey
Travel behaviour and barriers among non-users in the TNW fare network.
Nov
mobil.TUM 2025, Singapore
User behaviour in the dock-based and the free-floating system.
Dec
WWZ Economics Lunch Seminar, Basel
What drives e-bike sharing usage intensity.
2026
Mar
Submission to Transportation Research Part D
Article on the drivers of usage intensity.
Jun
Second advisory board meeting
Survey results, models and policy implications.
Aug
Speed-setting pilot extended
The initial two-year period ends and ASTRA allows the trial to continue.
Sep
Article in Transportation Research Part D
Volume 161, 105621, open access.
Sep
Article in the Journal of Urban Mobility
Volume 10, 100288, open access.
Sep
Thredbo 19, Valletta
Results of the speed-setting analysis presented at the international conference on competition and ownership in land passenger transport.
Q4
Final report to the SFOE
Findings, methods and policy implications of the whole project.
2027
Jan
Project end
Final advisory board meeting, archiving of the anonymised data sets and handover to partners, including the U-Two project.

Partners and advisory board

Funding

Swiss Federal Office of Energy, Energy Research and Cleantech programme, grant SI/502720-01 (ARAMIS)

Lead institution

Institute of Tourism and Mobility, Competence Center Mobility

Research partner

Faculty of Business and Economics (WWZ), Environmental Economics

Implementation partner

Free-floating S-pedelec sharing. Trip data and survey distribution.

Implementation partner

Dock-based pedelec sharing. Trip data and survey distribution.

Sister project

FHNW field test on fare integration of shared bikes with HSLU, BVB, BLT and TNW, 2026

Advisory board

  • WWZ, University of Basel (research)
  • Office of Mobility, Canton of Basel-Stadt
  • Basler Verkehrs-Betriebe (BVB)
  • Tarifverbund Nordwestschweiz (TNW)
  • Swiss Federal Railways (SBB)
  • Federal Roads Office (FEDRO)
  • Pro Velo beider Basel

Team

Portrait of Dr. Michael Stiebe

Dr. Michael Stiebe

Project manager, HSLU
Lecturer
Institute of Tourism and Mobility (ITM), Competence Center Mobility
Lucerne School of Business
Walter-von-Moos-Promenade 1, 2nd floor
6005 Lucerne
+41 41 349 36 24
michael.stiebe@hslu.ch
HSLU profile
Portrait of Prof. Dr. Widar von Arx

Prof. Dr. Widar von Arx

Co-project manager, HSLU
Professor, Head of the Competence Center Mobility
Institute of Tourism and Mobility (ITM)
Lucerne School of Business
Walter-von-Moos-Promenade 1, 2nd floor
6005 Lucerne
+41 41 228 42 88
HSLU profile
Portrait of Prof. Dr. Frank Christian Krysiak

Prof. Dr. Frank C. Krysiak

Co-project manager, University of Basel
Environmental Economics, WWZ
Room 4.51
Peter Merian-Weg 6, Postfach
4002 Basel
+41 61 207 33 60
frank.krysiak@unibas.ch
University profile
Portrait of Benjamin Weggelaar

Benjamin Weggelaar

Research associate, University of Basel, since October 2024
PhD candidate, Environmental Economics, WWZ
Room 4.47
Peter Merian-Weg 6
4052 Basel
+41 61 207 24 56
benjamin.weggelaar@unibas.ch
University profile