#
 

Parking Compliance Powered by Artificial Intelligence

June 20, 2019

Arnaud Dessein and Alix Paricard

How can we encourage compliance with on-street parking rules to make better use of public space and smooth traffic flow? Parking violations are a common problem worldwide and most major cities see their urban centers undesirably crowded. Cultural evolutions, new technologies, coupled with modifications in legislative frameworks, suggest an improved management approach for parking. How can artificial intelligence motivate a change in behaviors rather than repression of choice? One of the answers is ParkPredict Control: a tool that uses data to optimize on-street parking enforcement.


 


Organizational specificities in Europe and North America


Some European countries, such as Belgium, the Netherlands, the United Kingdom, Portugal, Spain, Norway, and France have reformed on-street parking regulations toward the decriminalization of parking enforcement and decentralization of all associated functions to local authorities. Once reserved for public agents, tickets can now also be issued by certified third parties like private companies. Such measures have generally been beneficial in terms of compliance with parking rules and improved spontaneous payment. For example, the direct payment rate increased from 35 to 90 percent in the United Kingdom and from 30 to 90 percent in Spain following the reforms. 


Another observed consequence is the reduction of road congestion by limiting the presence of overstaying vehicles, thus promoting access to city centers, reducing traffic and improving air quality. Finally, the financial gains of optimized enforcement can be reinvested in more sustainable and public mobility infrastructures such as public transport, alternative transportation modes or park and ride facilities.


In the United States, parking enforcement can be carried out both by the municipal police and private companies. Some cities also apply innovative measures, such as adaptive parking fares. In San Francisco, rates are adjusted per block and time period, and in Seattle, fees are based on the demand per price zone.


Farther north, in Canada, some cities such as Calgary and Victoria delegate parking enforcement to private organizations for efficiency issues. Calgary further calculates responsive parking prices based on the difference between the observed demand and occupancy targets. This again illustrates the importance of technology for optimized parking management and enforcement depending on the local context. 


A common insight from the various experiences of parking management in the world is that the organization of enforcement is a major lever of efficiency. In this context, making parking violations an administrative matter allows the adaptation of all enforcement issues according to local constraints, that a centralized system would otherwise neglect.


Another recurrent lesson to optimize enforcement functions is that the choice of technologies used is of utmost importance. In particular, enforcement can be facilitated with communicating systems for dematerialized payment and enforcement. For instance, pay-by-plate mechanisms coupled with automatic number-plate recognition (ANPR) technologies for enforcement, are ideal to leverage efficient operations.


ParkPredict Control is a decision support system to optimize enforcement rounds and on-street parking compliance. A major problem for enforcement teams is to know where and when to patrol in order to plan effective rounds.


Existing alternatives are based on traditional statistics like averages per hour, day or zone. However, this approach focuses on a simple historical analysis of the data taken out of context and out of the dynamic environment of the city. Moreover, the amount of data required is prohibitive to work finely on logistic plannings per street and per hour. Finally, these methods are often carried out by hand, which makes the work tedious, not replicable, and subject to miscalculation and misinterpretation.


This technological solution consists of enriching enforcement data by adding a large number of contextual data streams.


ParkPredict Control uses machine learning algorithms to predict where and when the number of violations is highest, and thus prescribe areas to visit. The machine is therefore able to determine the ideal route to optimize parking enforcement. A major advantage is the ability to accurately forecast and extrapolate the number of violations from a reasonable amount of historical observations. Moreover, the respective impacts of temporal and spatial variables can easily be decoupled, for recommendations by sector and period that are reliable, actionable and operational on the field. 


 


Data enhancement


This technological solution consists of enriching enforcement data (positions and times of checks performed and tickets issued) by adding a large number of contextual data streams. These contextual data describe the city as realistically as possible according to the elements impacting usages and behaviors (for example: the number, location and type of parking spaces, payment data by parking meter or mobile phone, the position and occupancy of off-street car parks, points of interest such as restaurants, schools or shops, weather, calendar data such as the hour, day of week, bank and school holidays).


After contextual modeling, the machine learning calculation engines work in two steps: calibration and prediction. During calibration, the models first train to learn the relationships between behaviors and context on all data collected in history. For example, the model assimilates that there are more violations on Fridays at 6 pm in a particular part of the city because many bars and restaurants are present. Once calibrated, models are able to predict the number of violations for a given location, since they have already observed similar contexts in the past.


Each new check also generates data that is injected into our predictive models to verify and improve their reliability. Thus, the models are recalibrated periodically, adding new data collected to the existing data. Predictive models can thus increase their performance and adjust to changes in behavior. Data related to urban infrastructure or socio-economic variables are also updated regularly to adapt to city structural changes. 


The recommendations are then made available to team managers in a dashboard and to agents on foot or drivers of ANPR vehicles via a mobile application. Violations against paid, reserved, obstructive, illegal and dangerous parking may also be included.


ParkPredict Control is thus an all-in-one tool for effective on-street parking enforcement powered by artificial intelligence. It becomes easier to fight against issues caused by violations and especially against urban congestion.


Arnaud Dessein is Data Scientist and Alix Paricard is Account Executive at qucit, based in Bordeaux, France. The company also has offices in New York City. Arnaud can be reached at  arnaud.dessein@qucit.comand Alix can be reached at alixparicard@qucit.com



#