Geofencing to Accelerate Digital Transitions in Cities: Experiences and Findings from the GeoSence Project

Societies are faced with many challenges that will require collaboration and innovative solutions to be solved. One of the challenges facing cities is the management - and planning of traffic, balancing between increased accessibility and availability for users while dealing with issues such as congestion, traffic accidents & incidents, and health and environmental issues. Investing in new physical infrastructure is needed but is usually very expensive and cities need to find more cost-effective ways to plan and manage traffic. New digital and connected solutions, such as geofencing, are being developed with the potential to effectively reduce these issues. Geofencing and how it is applied is shown in Fig. 1 below. Simplified, geofencing applications are based on 5 steps:

Bild vergrößern

GeoSence is a JPI Urban Europe project, funded by Horizon 2020, with the objective to design, trial and evaluate new geofencing concepts and solutions in cities and to propose new ways to deploy different geofencing applications. It started in 2021 and will finish in 2024 engaging partners representing research institutes, universities, and local/national public authorities from four European countries. Three cities: Stockholm, Munich and Gothenburg are involved in the project, focusing mainly on the following:

Complementary surveys, interviews and literature studies were conducted in the use cases. Also, surveys and interviews were conducted in 2021/2022 on the challenges and needs of European cities in using geofencing for urban traffic management [1].

There is no legislation that directly prohibits geofencing. On the other hand, there is also no legislation that explicitly allows it, although there are exceptions in some countries. This means that different actors need to find their own way. However, there are rules that affect the chosen solution, such as GDPR (EU 2016/679). The GDPR governs how personal data is allowed to be collected and used, which can be particularly challenging for a city. The city often does not want personal data but only aggregated data from users. This means that the chosen geofencing solution needs to meet this need. The city may also want to procure vehicles or transportation services that are geofenced. One lesson from the GeoSence project is that the technology is not yet ready for large-scale procurement, but that more pilots are needed regarding this matter. Currently, it is possible to procure solutions based on retrofitting of vehicles from 3^rd parties, but these solutions so far require personalization of vehicles for the technology to work.

Access to and harmonization of data is a critical prerequisite for new applications and innovations in geofencing to emerge in the city. However, harmonized data and data availability alone are not enough for geofencing to significantly influence the city's transport and mobility ecosystem. In an analysis of the city of Stockholm, we see that another key factor is the city's ability and capacity for transformation in this area. This transformation ability is a combination of data management-, digitalisation-, and innovation capacity, combined with the city's ability to work horizontally and mitigate internal lock-ins while improving collaboration with external actors. By exploring geofencing as a technology, we see that the city can make significant progress in preparing for the sustainable mobility of the future by enhancing its ability to apply innovative geofencing solutions. If the city is in parallel and at the same pace are strengthens its overall transformation capacity.

Implementing geofencing requires a technical device to accurately measure a vehicle's spatial position and a digital system to process the geofences and geospatial information. Position data for geofencing is typically collected via Global Navigation Satellite System (GNSS). The spatial accuracy of modern GNSS sensors is about 1–3 m under ideal conditions but it can be affected by various factors (e.g., tall buildings, tunnels, or dense urban environments), introducing position errors of several meters.

The use cases in Gothenburg and Munich illustrated the challenges posed by GNSS inaccuracies. In Gothenburg inaccuracy led to issues with the detection performance for a geofencing-based retrofitted Intelligent Speed Assistance (ISA) system. When two roads ran closely alongside each other - one within and the other outside a geofenced area - the system sometimes triggered erroneous feedback (i.e., false positive or false negative errors, both resulting in wrong speed limit suggestions). Even though this occurred rarely, it affected the perception of effectiveness and acceptance among the participating drivers negatively. However, the collected data offers the chance to identify and analyse such issues and to adjust and improve geofences and system functions.

In Munich, parking zones for e-scooters were geofenced. The inherent limitations of GNSS accuracy together with the small dimension of the parking zones (2 x 10 m) cause a low reliability to detect an e-scooter within a parking zone. To prevent this, a tolerance range (approx. 20 m) is therefore applied to verify correct parking. Nevertheless, users experienced that e-scooters could not be returned despite standing within a parking zone, and in other cases e-scooters were returned in non-parking areas outside a parking zone. Presently, GNSS accuracy often falls short of providing reliable data to regulate e-scooter parking in urban settings. Therefore, additional sensor technologies are currently being tested worldwide to enhance the accuracy of the position signal.

The second prerequisite is a technical solution to process position and geofence information to generate feedback for drivers and/or to enable direct vehicle control functions (e.g., speed regulation or permitting/rejecting e-scooter parking). This can be achieved either by an on-board digital device using pre-installed data (e.g., maps and geofences) or by sending the data via wireless broadband communication to a cloud-based service for processing and subsequent feedback information. The former approach demands more technically complex systems but provides more or less real-time performance. The latter method is suitable when temporal delays of several seconds between data transmission and reception are not critical for performance.

A prevailing challenge for geofencing implementations is the absence of standardized application interfaces and that few geofence solutions are available in new cars. Existing solutions are often closed systems with proprietary OEM-specific APIs for fixed functionalities. Retrofitted hardware is currently the only viable option for implementing geofencing solutions in a heterogeneous vehicle fleet, but applicability may vary. Barriers include elevated costs for retrofitted equipment and expenses for equipment installation, mobile data usage, and third-party services as well as potential issues concerning insurance and warranty rights.

Digital platforms are required to implement geofencing solutions in a meaningful way within an authority. The digital platforms are used to create and manage geofences and their policies in an easy and seamless manner, integrating it into existing digital processes and workflows, and processing data that enables the analysis, control and further adaptation of geofenced-based regulations (data for governance).

Our work in GeoSence revealed two ways in which authorities find appropriate digital platforms. One reported approach is that authorities develop digital solutions on their own, relying on open (source) and partly self-developed software, standards or data interfaces. The solution in this particular case was initially developed in an innovation project by a municipality in Norway and later used by other cities in the country. It provided digital maps with different layers of information for e-scooter companies including regulations for speed, access, parking and vehicle fleet size restrictions.

However, most municipalities will probably rely on procuring a commercial digital platform solution through a public tender. Especially for the micro-mobility sector (but also in the transport management sector) a number of commercial solutions have been developed in the past years. A commercial solution for geofencing in the micro-mobility sector has been implemented by Munich’s mobility department. The solution allows the definition and management of geofences and their policies and provides functionalities to transfer the rules to the e-scooter providers in a standardized way. The dashboard solution also retrieves data about e-scooter usage and parking from the providers, can display them together with rules on a map, and integrates various tools for the analysis of the data. As solutions get more developed and mature with time this approach to integrate digital platforms is perhaps a more appropriate way for authorities, since it will require less expertise and human resources to handle technical and IT-related issues in the long run. Further important decision criteria for selecting a digital platform are scalability and interoperability to fit the needs of new or higher demands in future.

Learnings from the interviews and surveys conducted with transport and traffic experts representing the public sector, in different European countries [2] are in line with the learnings from the GeoSence use cases. Some of the key learnings are:

(1) Potential, risks and barriers – One of the potential things mentioned are a need of fewer signs and maintenance, leading to more efficient traffic regulation. One risk mentioned is that transport companies will use geofencing solutions mainly to buy one's way of access in the transport system. One barrier mentioned is the limited economic resources in municipalities to make this transition. (2) Acceptance and stakeholder involvement – Acceptance of geofencing for traffic management is easier with stepwise implementation for shared types of transport, starting with smaller pilots before scaling up. Successful implementation also requires multi-stakeholder collaboration, including private entities, the public sector, research, and the users. For instance, it is important to involve the right stakeholder with the required competencies at the right time, e.g., to set geofencing zones. It is also crucial to start with the user needs, rather than with the technological solution. (3) Agreements and regulations – Voluntary agreements between e.g., municipalities and mobility operators are one way to implement geofencing solutions, and if done correctly can develop the relation between the actors. The problem is that with voluntary agreements, the municipalities cannot guarantee that the operator will comply with the proposed measures. In some cases, new transport modes such as e-scooters can benefit from clearer regulations making it both clear for the operator what they need to oblige with, and also making the space of enforcement for the municipality clear e.g., to what degree they can demand that operators use solutions, such as geofencing. (4) Data – It is important that there is access to desired data, that it is of good quality, that the data is updated frequently and that the data format used is interoperable and standardised. In some cases, municipalities, are setting their barriers own by not providing access to the desired data. Data sharing platforms for sending and receiving data make setting up and communicating the zones more efficient.