- Data insights
The commuter train network in Stockholm, Sweden generates a lot of data. Can large
amounts of traffic give commuters better traffic information and maybe even improve punctuality?
- Innovation
Using historical data we devised an algorithm (The Train Brain) that predicts delays hours before they happen. By monitoring real time arrival and departure times, the forecast model knows when a train is late and calculates the risk of further delays across the entire rail network. When trains are late The Train Brain predicts correct departure times 20 times more often than exiting traffic information systems.
The Train Brain is also implemented in a smart phone app, helping commuters avoid
potential future delays 2 hours in advance. It displays the position of every train in the system as well as showing real time seat availability in every train car.
- Originality
There are apps showing real time information in public transportation, but this is the first time this data is processed to forecast delays. Every 60 seconds The Train Brain recalibrates its forecast of future train departures helping commuters regain control of their daily commute while simultaneously improving punctuality.
- Data enhanced consumer experience
Since The Train Brain predicts correct departure times 20 times more often than previous traffic information systems, it increases the traffic control centre's ability to improve punctuality. This may be important benefit to consumers –arriving on schedule is what commuters demand.
But all delays cannot be avoided. The Train Brain gives access to more correct traffic
information and helps the commuters to take control of the situation.
- Data driven behavior change
15% of commuters in Stockholm downloaded in the first week after launch. It gives advance notice of future delays, helping commuters them plan their journey better. By reducing the number of people arriving at a busy station the algorithm has the potential to prevent delays caused by overcrowding.
- Business impact
Our client has incentives (bonuses as well as fines) based on a customer satisfaction index.
Customer satisfaction jumped by 30% among commuters who use the app or have heard of the Train Brain Project.
What started as an innovation project for the Stockholm commuter rail network can now be licensed as a tool for a punctuality revolution. Not only for commuter trains, but for people using public transportation all over the world.
A negative spiral of bad press concerning the Stockholm commuter train-network led to decreased customer satisfaction. By using all the different data the trains produce, we managed to create an algorithm that forecasts and sometimes prevent delays all together.
The data was converted and made available to all commuters in an app. A part from improving the actual quality of the service, the Train Brain algorithm created positive press with a global reach of over 500 million people.
Most importantly, the customer satisfaction increased with 30% among those using the app or that have heard of the Train Brain Project.
- Data gathering
For punctual traffic and for safety reasons railway traffic control gathers data of where every train is at any given second. This data was available to us through an API.
- Data interpretation
By monitoring live arrival and departure times, The Train Brain knows when a train is late and forecasts the risk of further delays across the network.
The model is similar to an earthquake seismograph, it looks for significant peaks. Trains pulling into a station behind schedule represent a peak. The algorithm automatically compares how a similar delay on the same line impacted the network on previous occasions, to predict how it will affect the network in the near future. And like a human brain it evolves. It learns and improves its ability to make forecasts over time.
This makes it possible to forecast service disruptions and traffic control can prevent the ripple effects that cause most of the delays.
- Targeting
The Train Brain is targeted to traffic control centres as well as commuters. It helps traffic control to improve punctuality. And its unique departure time predictions enable commuters to reschedule or choose an alternative route if the service is likely to be disrupted.