Trellint: Driven by data

How data science can support social equity in parking enforcement

Technology and its applications in the parking industry have ramped up in many different areas over recent years, with data becoming an integral part of any enforcement strategy.

Trellint, created following Modaxo’s acquisition of Conduent’s kerbside management and public safety businesses earlier this year, is leading the sector in its use of tech-led enforcement and data science, particularly for social equity in parking enforcement.

In the United States Trellint operates at scale supporting large cities like Los Angeles, Boston and Chicago, and there are many examples where the application of data science has helped to prevent disproportionate and higher fines for disadvantaged communities and vulnerable debtors. Where outdated enforcement schedules and routing allocations are still in place, data science gives cities a more targeted way of improving existing parking policies and deciding new ones.

The use of data science for parking strategies in the UK is still in its nascent phase but it is growing. Trellint is bringing its experience in the US to its local authority clients in the UK, many of whom understand how data can help drive decision-making but often lack the time and expertise to analyse it effectively.

Matt Darst, Trellint’s Director of Professional Services based in the US, said: “When looking at the parking sector in the US and the UK we recognise that there are some differences in approach. CCTV cameras are more widely accepted in the UK than in the US, for example, and GDPR provides a level of data protection in the UK which inevitably makes the public feel more comfortable about that use of camera technology. So, there is a cultural aspect to the approach we take in different geographical locations but the overall approach to customer segmentation and support remains similar at a high level.”

Application of data science

Local authorities having data analysts within their teams in the UK is becoming more common, but data science takes the use of data to the next level. Trellint’s data scientists have advanced degrees in machine learning, electrical engineering, and data science, as well as training in computer vision and AI.

Traditional data visualisations, while useful, don’t provide essential information on the complex relationships between data sets, and often don’t identify the hidden trends behind the data or help predict the outcomes of any given scenarios. Visualisations created by data analysis can help tell a story of a community, but the more in-depth work of deciphering data might then become the burden of city officials who don’t have the time or expertise to scrutinise it and make recommendations. This is where data science can help.

Access to data varies from city or council, and how you solve problems using data varies. If Trellint builds an algorithm for one city, that model cannot simply be copied and used elsewhere as the inputs will be different. Trellint’s data science teams consult with municipal clients to help them improve operations and address the variety of challenges they face, which could include congestion, safety and increasingly equity in parking enforcement.

Matt Darst explains: “Many of the parking policies we see are decades old, and our clients and the whole sector generally, hasn’t really taken advantage of this new approach to using data to relook at policies that have been put in place many years ago, with many cities still using the same enforcement schedules or enforcement routes used decades ago. They neither have the time to dig into the data – as city resources are overburdened as it is – nor the required big data expertise.

“Our data science team is both highly qualified and focused exclusively on kerbside management, allowing us to spend less time on data wrangling and move faster to make recommendations, including kerbside policies that don’t disproportionately affect disadvantaged communities.”

Why equity matters

Trellint believes that outdated parking assumptions can lead to entrenched community problems. When creating an enforcement policy, every component of that process is open to reinterpretation and reengineering to improve how fines and kerbside access impacts those communities.

Trellint works to identify opportunities for improvement so that there is no disproportionate fining of vulnerable debtors. In the US they have found that fines are often unjustly frequent, and also higher due to the way fines are structured.

The unintended consequences of inadequate planning and relying on historical plans, is that parking enforcement isn’t optimised. Despite the innovation in the industry in recent years, the effective use of the kerb to reduce congestion, improve safety or promote equity still hasn’t been achieved.

Using data science, cities can be mapped to show the likelihood of infractions in certain zones, including the times when they are most likely to occur. This allows the local authority to deploy personnel correctly and ensures that they work more efficiently within new enforcement zones, focusing on the areas with the most serious infractions which in turn improves equity.

Matt Darst expands: “Trellint’s work in Chicago is a great example of data science in practice. We increased the number of enforcement zones from 46 to 192 and prioritised those zones based on the likelihood of dangerous or congestion-related infractions. We then created the
schedules for the city. Since we’ve been doing that, we’ve seen a real shift in PCN issuance, from neighbourhoods that are at risk to neighbourhoods where parking concerns are much more serious.

“Where an infraction doesn’t cause a dangerous condition or create serious social harm, e.g., an expired licence plate, the level of fine should be lower to reflect the impact. Further, the impact to vulnerable motorists often compounds. Disadvantaged motorists are often less likely to contest PCNs, and more likely to be found liable. There’s a correlation between literacy rates, incomes and the likelihood of being found liable. Our use of data has led to us making recommendations – like implicit bias training and the use of visual cues, like images and symbols – which make adjudication fairer.”

Clamping vehicles in underserved communities can also have negative consequences as those are the people that tend to need their cars to get to work, school and elsewhere. Denying them access to their vehicles can create a cycle of debt.

“Chicago realigned its debt collection policies, focusing less on disadvantaged communities based on data we provided,” says Matt. “This has led to a higher number of clamps being deployed in less disadvantaged communities and increased productivity due to the reallocation of personnel.”

The US is ahead of the UK in terms of using data scientists for bespoke strategies and a review of the UK’s two-tiered enforcement system is due. While the cost of employing full-time data scientists can be prohibitive for councils, consultancies can provide a proportion of the necessary work but Trellint provides the full data consultation and implementation service, making it unique in the UK market.

Collection of data for kerbside management can now be achieved through mobile registration plate capture technology – KERBSight – which works alongside civil
enforcement officers, allowing them to cover a wider area.

Matt concludes: “It’s important to build feedback loops to ensure that recommendations remain fresh and relevant so that they align with reality. Today’s data-driven insights are tomorrow’s assumptions, so you must constantly review data to ensure you’re on a continuous path of improvement.”

Trellint’s KERBSight

Designed to maximise the potential of customer operations through data-driven insights, Trellint’s KERBSight provides customers with comprehensive access to data on the availability and usage of both on-street and off-street parking spaces.

A self-contained camera is used on an electric scooter, with its own power source, so as not to take any battery power from the bike. It is equipped with an in-built management system, featuring ANPR cameras and embedded GPS technology. This bike patrols designated areas, capturing vehicle registration marks (VRMs) and transmitting the data to the KERBSight management suite for analysis.

Through this data, the system provides valuable compliance data, such as the validity of permits, cashless session status, and Pay and Display (P&D) session information. This integrated approach enables kerb management of the kerb and optimised operations.

Additionally, Trellint can also offer detailed insights into various aspects of vehicle data, including make, colour, tax status, MOT status, year of manufacture, engine capacity, CO2 emissions, and fuel type, which can support business cases for future projects, such as a new controlled parking zone.

KERBSight is back-office agnostic, meaning you do not need to use Si-Dem, Trellint’s propriety back-office enforcement processing system, to benefit from the technology, Trellint just needs access to parking session and permit data from existing providers to maximise the benefits offered.

Matt Darst is Trellint’s director of professional services based in the US
trellint.com