In 2016, Montréal’s Mobility Operation Center Division (DEIGD) became conscious of the rapidly growing complexity of data management and system interface definitions while completing pilot projects on live traffic data collection. Data is often compartmentalized in separated applications, inaccessible broadly and timely for direct consultation. The DEIGD was looking for a new architecture to integrate various data sources and enable the required scalability. This data is produced by equipment (traffic light controllers, counters & detectors, LED displays) installed and maintained by the city. Many systems are considered safe or mission critical, most outdated technologically. The primary need for the DEIGD is to maintain and expand the systems over an extensive period (more than 25 years). To achieve this, we need to:

• Maintain a high level of consciousness toward technical debt (for durability and evolutivity)
• Implement, preferably, maintainable, durable and highly scalable systems.

The DEIGD wanted to answer this fundamental question: How do we prepare for tomorrow’s technologies without knowing the orientation’s outcome? Subjects like vehicle automation and electrification will challenge public organization expertise seriously. How do we prepare for this in an environment that hasn’t changed in 50 years? The conclusions of the various studies and pilot projects identified the following needs:

• To implement a central data distribution system and manage “data dictionaries” between data subscribers.
• To develop a data centric distribution system with interfaces only described with data and meta-data.

• To implement a Data Distribution Service (DDS™) described by OMG as a core engine to “forge” data at the ingestion level.

• To develop a system’s architectural “tricks” and ingest any data, enabling the system’s interoperability and evolutivity.

In conclusion, the implemented architecture completely decoupled the data from the systems. This data decoupling is fundamental and promotes interoperability at its highest possible level. The DEIGD implemented this architecture in 2021 and is now ready to integrate most of its data collection and is aiming for complete data migration by the end of 2024. The presentation will explain the findings and conclusion of a journey of more than five years to develop the latest data centric distribution system architecture, thus enabling massive possibilities for futuristic data integration.

François Thibodeau graduated from the École de Technologie Supérieure in Montreal in electrical engineering (1997). He now works as a systems engineer. He is recognized for his expertise in the design and integration of real-time systems pour la division de l’exploitation, de l’innovation et de la gestion des déplacements à la ville de Montréal (DEIGD). He develops the tools and architectural concepts necessary for the technological positioning required of municipal organizations to face the challenges of vehicle electrification and automation. François promotes the city’s open architecture policies in connection with the data-centric approach, thus allowing interoperability and control of system governance. He promotes an approach based on maintainability and the integrity of the data flowing between systems, thus ensuring the sustainability and scalability of the systems.