**6. Conclusions**

This paper presented practical considerations from the development of a parking occupancy simulator. The simulator helped in assessing suitability of an SPS suitability for a university campus, and its design and development contributed to reduce the SPS development efforts. The simulator shares software components, GIS data, and services with the SPS. The paper commented on design decisions made regarding the SPS and the parking occupancy simulator, which may be used in similar contexts to minimize development efforts. The application of the proposed methodology on the simulator development will produce software components readily usable in an SPS, as well as GIS data and services readily available online when published on a GIS sever. The software created for this work is available in a public repository, for further technical details inspection and to encourage reproducibility. The paper also presented experimental evaluation of SPS usage benefits using the simulator. The experiments differentiated profiles of agents that use the SPS and those who decline to use it. Analysis of experimental results showed how to use the total driving distance as a metric for evaluating the SPS benefits from a driver point of view. The experimental results also allowed us to explore the effects of having a parking reservation that is logical but not physical, clearly showing that the drivers who decline to use the SPS may "steal" a significant number of already reserved parking spot, which in our experiments reached numbers higher than 100 for a parking demands above the 40 percent of the maximum demand. The guarantee problem becomes more significant as the level of usage of the SPS increases, topping out at 50%, which may hamper drivers' acceptability towards an SPS.

The parking choices used in this work are simple, as they were not among its main goals. Future improvement directions could include the incorporation of more realistic choices, driving behavior, and the ability to learn from experience, which may be combined with an improved reservation algorithm in order to provide agents with updates upon changes in the occupancy state, e.g., when other alternative spots become available. Additionally, a comprehensive model validation with a study case for a different scenario and parking occupancy measurements taken in the field is planned as future work. The design and development guides and experimental analysis presented in this paper show the convenience of including simulations when considering the application of an SPS to a scenario, while diminishing possible dissuasive aspects like model complexity or simulation development efforts, that SPS development teams may consider.

**Author Contributions:** Conceptualization: Germán M. Mendoza-Silva and Raul Montoliu; Methodology: Germán M. Mendoza-Silva, Michael Gould and Joaquín Torres-Sospedra; Investigation: Germán M. Mendoza-Silva; Software, Germán M. Mendoza-Silva; Supervision: Raul Montoliu andMichael Gould; Resources, Joaquín Torres-Sospedra and Joaquín Huerta; Writing—original draft, Germán M. Mendoza-Silva and Michael Gould; Writing—review & editing, Michael Gould, Joaquín Torres-Sospedra and Joaquín Huerta

**Funding:** We thank funding from the Spanish' Ministerio de Economía y Competitividad under the project 'SmartWays' (Convocatoria Retos-Colaboración, RTC-2014-1466-4). Germán M. Mendoza-Silva gratefully acknowledges funding from gran<sup>t</sup> PREDOC/2016/55 by Universitat Jaume I.

**Acknowledgments:** We thank our university's Management Department for providing campus usage and community information.

**Conflicts of Interest:** No potential conflict of interest was reported by the authors. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
