Case study: using artificial intelligence in the zero-emission transportation sector

Artificial intelligence serving the user experience in public transit

Case study: Alstom and exo

Faced with the rapid transformation of digital technologies, public transport operators are rethinking how they interact with users.
Artificial intelligence (AI) is now emerging as a central lever to provide a smoother, more predictive and personalized mobility experience.

This case study presents two concrete examples:

• Exo, a public organization that operates train, bus and paratransit services in the territory of the Montreal Metropolitan Community, which developed the systemChrono SAEIV, an operations and passenger information platform integrating AI to predict punctuality and inform users in real time.
• Alstom, a global player in sustainable rail mobility, which uses AI to analyze passenger behavior, optimize onboard comfort and anticipate their needs.

These two complementary approaches demonstrate how data and artificial intelligence can transform the relationship between operators and their users.

Context and needs

In a context of accelerated digital transformation,exo and Alstom illustrate two complementary approaches to using artificial intelligence for the benefit of public transport users.

Forexo, the main challenge is to improve thereliability, transparency and punctuality of the network at the metropolitan scale of the greater Montreal region. Inheriting a multiplicity of systems and local networks, the organization had to deal with complex management of schedules, incidents and passenger information. Its objective is twofold: to providereliable and predictive real-time information in order to strengthen user confidence, while developing asustainable and scalable, capable of integrating future technological innovations to optimize operations and service planning.

For its part,Alstom is seeking to adopt a more global approach integratingthe dimensions of comfort, safety and service quality for passengers. The company decided to implement a continuous improvement approach to maintenance and operations. This aims to strengthen the reliability, safety and overall performance of railway networks, while using collected data to anticipate needs and adapt services to passenger expectations. Alstom's ambition is toput technology at the service of people, relying on artificial intelligence to offer a smoother, more intuitive and secure mobility experience, while guaranteeing thethe protection of passenger data.

Artificial intelligence solutions implemented

To meet the needs of users,exo andAlstom have relied on AI solutions that combine technical performance and tangible improvements to the customer experience, exo positioning itself as a true pioneer in this area, well before AI became a mainstream topic.

Atexo, theOperational Assistance and Passenger Information System (SAEIV) was developed internally and has now established itself as a central platform for metropolitan mobility management. Originally designed by the Agence métropolitaine de transport (AMT) to provide real-time tracking of trains, the system has progressively evolved to also integratebuses and passenger information. It is based on amachine learning model leveraging between3 to 5 billion data points per year, allowing it topredict the punctuality of vehicles and anticipate disruptions related to weather, construction or traffic. Thanks to anautomatic incident detection and to aproactive communication via signage, SMS, social media and public APIs, exo offers reliable and dynamic information to users. Hosted on acloud infrastructure ensuring an availability of99.95%, the system allows travelers toplan their trips with confidence and ease |||NEXT|||.For its part,

Alstom uses artificial intelligence to provide asmoother and safer passenger experience |||NEXT|||.By analyzingticketing and video surveillance data, its systems identify travel habits, crowding levels or passengers' preferences, allowing services to be adjusted accordingly. At the same time, aproactive monitoring detects incidents such as health issues, fire risks or abandoned luggage, reinforcing thesafety and confidence on board |||NEXT|||.Finally, the analysis of information from videos and ticketing collections also makes it possible to evaluate the average number of passengers per trip, to identify users' preferred locations as well as the use of storage spaces for luggage and bicycles, thereby contributing to a more nuanced approach to onboard comfort. This detailed knowledge enables optimization of onboard services, limiting overcrowding issues and improving overall comfort. Alstom's objective is clear:make journeys more comfortable, personalized and safe, while integrating technology in a responsible and humane way.Observed results

The results obtained by exo and Alstom concretely demonstrate the positive impact of artificial intelligence on service quality and the user experience.

For exo, the deployment of the Chrono SAEIV system resulted in more reliable and predictive passenger information, reaching an accuracy rate of 90% within a five-minute margin. Thanks to proactive communication during disruptions, whether due to construction, adverse weather or detours, passengers can now plan their trips with greater peace of mind. The tool also supports operational teams by facilitating planning, incident detection and continuous analysis of network performance, notably in terms of punctuality, ridership and vehicle availability. With more than 3 billion data processed each year and 1,400 vehicles tracked in real time, representing nearly 28% of Quebec's public transport fleet

[1], exo now has a high-performance and scalable platform. Result: a smoother travel experience and strengthened trust between the operator and its users.On Alstom's side, the benefits of AI translate into a better understanding of passenger behavior, thanks to the cross-referencing of actual usage data and observed preferences. The analysis of information from videos and ticketing collections also makes it possible to assess the average number of passengers per trip, identify passengers' preferred locations, and the use of storage spaces for luggage and bicycles, thereby contributing to a more refined approach to onboard comfort. This detailed knowledge enables optimization of onboard services, limiting overcrowding problems and improving overall comfort. Furthermore, proactive system monitoring contributes to increased safety by detecting anomalies or incidents early that could affect travel. Ultimately, these innovations make public rail networks safer, more reliable and more profitable, to the direct benefit of travelers.

Lessons and perspectives

The experiences conducted by

exo andAlstom offer several key lessons on the successful integration of artificial intelligence into public transport.First, the

coherence and explainability of models are essential conditions for success. In both cases, the AI solutions weredesigned from real on-the-ground needs,whether it is operational management for exo or maintenance and passenger comfort for Alstom, with an approach centered on thecreating value for the user rather than on technological performance alone.Next,
the quality and unification of data remain an indispensable foundation: without reliable, harmonized and well-structured data, no AI model can produce sustainable or credible results.A solid governance and a long-term vision, based on a strategy that integrates product governance, continuous data collection, regular model updates and anticipation of future product needs, are essential to ensure its longevity and coherent evolution over time. exo, for example, chose to develop its tools internally, thus guaranteeing the

long-term viability, transparency and control over technological evolutions, rather than adopting a solution subject to a contract with a vendor offering its own solution.Finally, trust and ethics

are at the heart of these approaches: theprotection of privacy andcybersecurity are not peripheral add-ons, but indispensable pillars for any AI deployment in the public sector. By combining responsible innovation, transparency and respect for users, exo and Alstom are models to follow of anartificial intelligence in the service of sustainable and human mobility.ConclusionThe cases of exo and Alstom demonstrate that artificial intelligence is not just a tool for operational performance: it becomes a

driver of human experience

.By transforming data into intelligence, these organizations create the conditions for public transportmore reliable, safer and closer to users' needs
, thus contributing to sustainable and connected mobility.AI in the service of ethicsHowever, artificial intelligence cannot be deployed without reflection on its

essential condition for its use

. There are resources to support transport stakeholders towards anethical, reliable and sustainable AI.
TheDeclaration of Montreal

• is addressed to policymakers as well as to any individual, civil society organization and company wishing to take part in the development of AI in a responsible manner. It lists principles that serve as the directions of an ethical compass to guide the development of artificial intelligence toward morally and socially desirable ends.Here is also aresource offered by the Order of Engineers of Quebec

• . This document presents a synthesis of the 6 areas of vigilance of the Professional Practice Guide, highlighting the recommendations of the Order of Engineers of Quebec on the responsible use of artificial intelligence.We encourage you to adopt a thoughtful and responsible approach in your use of artificial intelligence. The resources we provide below are not exhaustive; we therefore invite you to deepen your knowledge and remain attentive to best practices in ethical AI.. Ce document présente une synthèse des 6 axes de vigilance du Guide de pratique professionnelle, en soulignant les recommandations de l’Ordre des ingénieurs du Québec sur l’utilisation responsable de l’intelligence artificielle.

Nous vous encourageons à adopter une approche réfléchie et responsable dans votre utilisation de l’intelligence artificielle. Les ressources que nous vous proposons ci-dessous sont non exhaustives ; nous vous invitons donc à approfondir vos connaissances et à rester attentifs aux bonnes pratiques en matière d’IA éthique.