+44 (0) 15 64 793552

ETC Conference Papers 2017

Spatial impacts of autonomous vehicles compared to railways

Submitted by / Abstract owner
Robert Möhring

Robert Möhring, Delft University of Technology, Niels Van Oort, Delft University of Technology

Short abstract
Automated Vehicles (AVs) could develop into an individual form public transportation and weaken established collective transportation. Aim is to estimate how much space AVs will require for a fictional replacement of a given railway service.

“Automated Vehicles” (AVs) are proposed as promising new mode of transportation worldwide and along with increasing automation expectations rise towards vehicle capabilities. As an example, AVs are thought to enable less congestion, shorter headways and energy efficiency. AVs are tested in Sion, Lausanne or Masdar as well as in the Dutch environment of Ede-Wageningen.

The Netherlands are one of Europe’s smallest but most dense populated countries. Commuting between cities is a daily routine of many people and done by private vehicles or public trains. In the future, AVs could develop themselves into an individual form public transportation and thus, weaken established collective transportation such as commuter rail. On the other hand, AVs could strengthen public transport by enhancing the first and last mile.

A major concern within a small and dense country like the Netherlands is the efficient usage of urban and rural space. However, the effect of AVs on spatial issue’s is still not clear either and due to the density of the Netherlands, it can be discussed, that fully automated vehicles potentially replace national railways services in the future. Thus, replacing a given railways service by AVs will impact spatial usage and availability since it will also require space along a given or proposes (road) infrastructure.

In this paper, the focus is on the number of AV needed and space those AVs require in order to replace a given railway service of the Netherlands. Spatial occupation of different AV types is estimated in relation to different passenger demands and the results are compared to the spatial occupation of an established railway service.

The number and spatial impact of different AV types (Automated Car, Masdar PRT, EasyMile EZ10, Singapore GRT, Navya Arma) is calculated by using variable passenger demands, variable vehicle speed and variable time headway. Subsequently, this value was compared with the spatial impact of the current fleet of intercity and regional trains from the Nederlandse Spoorwegen N.V. between The Hague Centraal and Utrecht Centraal.

The number-AV per type-AV was estimated by dividing various passenger demands by capacity per type-AV and multiplied with occupancy factor per type-AV.

The spatial occupation of the estimated number-AV per type-AV is calculated as follows:
1. Time headway is transformed into space headway (via m=v*s).
2. Estimated space headway (depending on speed) is added to vehicle length (depending on type-AV).
3. This summation of space headway and vehicle length is then multiplied with the vehicles’ width, resulting in spatial occupation per type-AV.
4. The occupied space per type AV is then multiplied with the estimated number-AV per type-AV, resulting in the spatial occupation of the estimated number-AV per type-AV.
The spatial occupation of the estimated number-AV per type-AV is then compared to the above mentioned special occupation of the train service. Calculations were made using Excel 2016 (Microsoft Office 2016, USA).

The faster AVs operate, the faster their spatial impact increases. With the objective to keep spatial occupation on the level of a given railway service, the number of passengers transported by AVs decreases with increasing speed.

As example: The investigated railway service has a spatial occupation of 4,420.65 m² and offers 1,610 pass/h transit (calculated with an occupation rate of 0.3). AV occupy the same amount of space but are capable to transport 100 to 140 pass/h (160-120 km/h respectively). Thus, a (public) transport system based on AV could transport up to 11 times less passengers than the investigated railway service, considering spatial occupation of both modes is equal.

Conclusion & Recommendation
The number of AVs needed to replace a railway service would require a huge fleet of vehicle at first, which themselves would occupy more space than a given train service. As a mode for mass transportation, AV’s space occupation would be highly inefficient and collective transport modes should therefore be prioritized. Associated to railway services, similar results are expected for MRT, LRT and BRT. In this context, the automation of trains shall further be investigated. However, AV offers potential for low occupied busses or public transportation in rural areas and therefore a potential efficient mode for first- and last-mile transportation.


No documents available.


Association For
European Transport

Forester House
Doctors Lane,
Henley-in-Arden B95 5AW,
United Kingdom

+44 (0) 15 64 793552


VAT number: 710 1866 64

Conference Supporters & Endorsers