Predictive occupancy for a return to safe travels.
Data-driven Public Transport
Read our free guide to data-driven public transport here.
The Hello World Podcast:
Meet Kim and Jan from Stratiteq and watch our interview with Birgit Wirth from Deutsche Bahn and Arriva UK Trains on traveling in the new normal era.
Relevant client cases:
Skånetrafiken became a digital pioneer.
Optimized Public Transport through Advanced Analytics.
Skånetrafiken is the third-largest public transport operator in Sweden with 170 million passengers per year, within Skåne county and in transit between Sweden and Denmark.
Public transport has been highly affected by Covid-19. Skånetrafiken is the operator in the southernmost part of Sweden, responsible to secure that citizens and visitors can travel to work, school and to places of interest in a safe and sustainable way. During the pandemic, the definition of fully occupied buses or trains changed overnight. This made Skånetrafiken look for new ways of providing safe traveling, and they knew the solution was going to be found in modern technology.
“We’re always looking to improve the experience for our travelers and are continuously in a process to find innovative ideas, supported by new technologies. Stratiteq is a long-term partner and with their knowledge of our industry and how to use data in pioneering ways, we have created several sustainable and future-proof solutions for our travelers. Predictive occupancy was based on Stratiteq’s ideas on how to help commuters travel safer in times of a pandemic”, says Johan Frithiof Karlberg, CIO Skånetrafiken
The solution was found in Skånetrafiken’s gathered data. Having a closer look, Stratiteq defined the ability to predict occupancy as a way of providing safer travelling. The solution was built upon journey search data, ticket purchases and ticket validations from boarding. Combining this data made it possible to determine both how many people will board at each stop, and predict where they will disembark. Hence, we can calculate how many people are onboard, compare that number with the actual vehicle capacity and calculate occupancy levels.
“This is without passenger counting solutions. If we add datasets like Wi-Fi information, smartphone telemetry, video cameras, weight sensors and other data sources, we can increase the granularity and predict, for instance, which car is more occupied than another in a train set. We have the technology and the data, and it’s time to understand and fully harvest the business value”, says Jan Landelius, Senior Consultant at Stratiteq.
Creating a data-centric architecture provides the capabilities not only to collect, store and analyze data, but also to build applications and services. Using Microsoft Azure as a data platform brought security, new technology, AI-powered services, real-time analysis and the capability to scale the solution as needed. The occupancy service is built on top of several Azure cloud capabilities with the core of the architecture being a data driven solution.
Development took advantage of the high degree of interoperability between the platform services. There’s also the low threshold for starting and setting up these components since they’re all Platform-As-A-Service (PaaS). The infrastructure and level configuration of the operating system is outsourced to Azure so the development team can focus on innovation and development.
The information was built into Skånetrafiken’s mobile app. Inside, the passenger can search for the journey they wish to make and based on historic data get a prediction of how busy the different departures will be. They can choose the ride they feel the most comfortable with from an occupancy perspective.
Predictive occupancy not only creates a valuable service to travelers. It also helps traffic planners to understand when additional buses or more train cars need to be deployed. This opens the door to better cost efficiency while maintaining a high degree of service.
New flexible tickets
The pandemic made many people stop commuting to work every day and periodical ticket purchases went down. By analyzing the data, a new, more flexible ticket was offered, keeping commuters from leaving public transport in favor for a car.
Using occupancy data makes it possible to plan operations more accurately – number of buses, size of buses and trains, driver schedules etc. which will increase efficiency and reduce cost. By adding weather data and traffic data it’s also possible to make quick adjustments and improve customer satisfaction as both availability and accuracy will improve.