Smarter cities need more than smart algorithms. They need trust

We’ve never had a better understanding of how cities move, but turning that knowledge into smoother, cleaner and more equitable trips remains a challenge. The main obstacle to sustainable mobility may not be technological, but rather political and organizational. Apps track buses in real time, sensors count cars and pedestrians and algorithms predict traffic jams. Advances in data-driven analytics even promise to let bus networks adjust service frequency to demand or to co-ordinate deliveries with fewer vehicles. And yet, getting across a city is often still a struggle: breakdowns, congestion, double-parked delivery vehicles blocking buses and cyclists, night shifts that end after the subway has stopped running. Improving urban mobility without increasing emissions or deepening inequalities in access means making better use of existing data rather than building more infrastructure. One thing is becoming clear: mobility data is still underutilized. It can only transform how we get around if it’s shareable, representative and actionable. An ambulance in a traffic jam. (Yvon Gallant/Pexels) So how can data become a real driver of sustainable transportation? That’s the question we explored with transportation and operations researchers from two Quebec research institutes (GERAD and CIRRELT) and the Canadian Operational Research Society (CORS), along with Québec-based mobility stakeholders, including at a symposium held last March at HEC Montréal. What data doesn’t show First and foremost, data must reflect reality. Research conducted in Toronto has shown that optimal delivery routes calculated using conventional data were often ignored by drivers, who knew an intersection was impassable or preferred a familiar route. When recommendations are designed to “make sense” to them, they are followed much more readily. But for recommendations to reflect reality, they need data that accurately describes how people actually travel. Sustainable mobility hinges more on the quality and sharing of data than on other anticipated innovations such as fare integration. Data flow (the seamless sharing of information among users, operators, and public agencies) is the central challenge of sustainable mobility. With this in mind, Québec mobility stakeholders are already taking action. For example, the Autorité régionale de transport métropolitain (ARTM) is preparing an open data hub modelled on London’s. In London, Transport for London’s open-data policy generates up to £130 million a year in value, according to a Deloitte analysis. Open data means more transit trips, time saved for users through real-time trip-planning apps and value created by the 600-plus apps built on this data, which more than 40 per sent of Londoners use. All with no additional public investment. But this openness runs up against a different logic in the private sector: once a platform aims for dominance, hoarding data becomes a competitive advantage, which stifles the entire ecosystem. Transit, a multi-modal app available in more than 1,000 cities, has seen this first-hand: many transit agencies want to open their data, but private players are holding things back. Three main challenges are hindering data sharing: the lack of common standards, inadequate user representation and information lost through anonymization. Roundtable with Hamzeh Alizadeh (ARTM), Anne Mercier (GIRO), Nicolas Fillion (AMD), and Sam Vermette (Transit). (Courtesy of the authors) Data that doesn’t speak the same language The first challenge is a practical one: there are often no common standards for collecting and formatting mobility data. As a result, one operator’s datasets are incompatible with another’s, and information stays siloed when it should be cross-referenced. In Montréal, parking regulations are a case in point: each borough manages its rules differently, making citywide standardization difficult. Without common standards, even the best data remains difficult to use: an algorithm trained on data from one network cannot be directly applied to another, and decisions made at the operator level cannot be optimized across the entire system. Interoperability — the ability of systems to exchange and use one another’s data — is essential if data-driven tools are to deliver real benefits citywide. What’s more, because data collection is often contracted out to private companies, transit agencies don’t always own their own data. Without a shared data repository and agreement among stakeholders, each entity optimizes within its own silo, and system-wide gains remain out of reach. Who is represented in the data? The second challenge concerns representation. Systems are designed for an “average” user who does not exist. And those who are least well served are often the least represented in the data used to design these systems. This isn’t just a statistical issue: these users are the ones who rely most heavily on public transit. More inclusive surveys would help make them visible. This is exactly what the the Autorité régionale de transport métropolitain (ARTM) is trying to address: identifying where the gap between supply and demand is widening, particularly in areas where vulnerable populations are concentrated. More representative data would, for example, help identify areas where poor service forces low-income households to own a car, when adjusting frequency or routes would be enough to eliminate that need. Improving service therefore requires understanding how people actually travel and what they need, and that starts with collecting data that reflects who really uses the network and who is excluded from it. Anonymizing without rendering data useless The third challenge pits privacy against utility. To protect users, data is anonymized, but indiscriminate anonymization can erase the very information that makes it valuable. For example, it becomes impossible to link an electric bus’s energy consumption to driving style in order to identify more efficient driving practices. Protecting users and serving them well can pull in opposite directions. Many studies show how important it is to know users’ profiles and preferences in order to tailor service. For example, research on car-sharing conducted with Montréal-based car-sharing company Communauto shows that if the service were to disappear, a significant proportion of its members would buy a car, leading to a direct increase in the number of private vehicles on the road. Communauto’s environmental benefit could only be quantified because users agreed to share their travel data. The same goes for electric vehicle charging stations: drivers vary greatly from one another, and a network designed as if they were all identical becomes overloaded in some areas while remaining under-utilized in others. Projected growth in car ownership without Communauto, from a study by Émilie Savard, Martin Trépanier and Catherine Morency. (Courtesy of the authors) The algorithm is not the obstacle These three obstacles have one thing in common: none are technical in nature. Meanwhile, pressure on roads shows no signs of letting up: with the boom in e-commerce, the number of delivery vehicles could grow by more than 60 per cent worldwide by 2030, and drone delivery is reaching cities faster than the rules governing it can be written. Digital technology isn’t what’s holding us back. It’s the systems of incentives, standards, trust and political will that determine whether better tools make for better cities. While research advances rapidly, the operational gap to implement it continues to grow. Bridging this distance requires governance, not just algorithms.