RATPPublic Transport Operator Upgrades its Location Intelligence with Prevision.io.
Prevision’s enterprise AI platform is accelerating and taking the friction out of our data science, AI and machine learning projects.
A leader in connected, sustainable urban mobility
RATP Group is a top-five player in the public transport sector globally. It operates and maintains passenger services in 13 countries on 4 continents.
The group has the charge of moving millions of people in major metro areas every day. It also operates a property portfolio of stations and transport hubs and has corresponding retail space that it needs to optimize.
Creating value from traffic flow data
With a wealth of ticketing, turnstile, and digital data relating to traffic flow at its disposal, RATP Group wanted to analyze movement patterns and geo-spatial information with a view to better understanding the evolving characteristics of each neighborhood.
In effect, this is essential to understand the service provision needs of different populations of commuters and travelers.
The groups also wanted to be able to improve how it commercializes its own available retail space. However, it also understood that there was a real opportunity to offer packaged location intelligence solutions directly to retailers at large: Armed with a better understanding of their catchment areas and foot traffic patterns they are able to make smarter decisions as to where to operate and what to sell.
Supporting business innovation with more agile data science
RATP Group was interested in putting its rail network data to use in innovative ways. However, there were only so many special projects that its data science team could take on at any given time.
For a project such as this, where there are many alternative approaches to explore, the group didn’t have the luxury of time to build and evaluate a multitude of different algorithms by hand to see whether they wanted to take the project further.
The data science team required a solution that would allow them to explore and validate different solutions very quickly.
Taking the friction out of machine learning
The data science teams at RATP Group were already well versed in data projects but were short of time. They thus appreciated the hands-on support of the teams at Prevision who were able to show them the shortest path to best leverage the platform’s capabilities.
Importantly, RATP Group’s specialists are able to unpack and therefore trust the performance of automatically generated models, which makes it much easier to gain buy-in internally.
Secondly, the ability to go much faster with automated machine learning is enabling RATP to take a test-and-learn approach to exploring innovative subjects. A team of just 2 was able to develop and evaluate 68 models in under 3 months – something that would have been technically impossible in the past. Furthermore,RATP Group has selected 6 of them to power planned apps.
Finally, Prevision is proving to be an effective sales tool: Because visualizing results is only one click away, it’s much easier to sell data projects to RATP Group’s external clients much earlier and before committing to productizing them.
“Prevision’s enterprise AI platform is accelerating and taking the friction out of our data science and machine learning projects,” said Paulin Assouvie, Program Manager for Strategy, Innovation and Digital at RATP Group.
Radical acceleration of AutoML is enabling a test & learn approach to data science projects
8 new AI-powered apps selected for production in under 3 months.
Greater visibility and faster time-to results with AI projects