Unparalleled gridlock forecasting app attributed to unique stochastic AI optimization algorithm
|Funding from Vinnova||SEK 500 000|
|Project duration||August 2018 - February 2019|
Purpose and goal
The project aimed to show the benefits of the XIMANTIS Forecaster on congestion hot spots, to proactively manage a journey for a positive environmental impact, improving efficiencies and quality of life. We devised a real-time data collection network, able to issue instant predictive analytics for vehicle navigation systems and traffic control centers (proactive traffic lights). We were also able to greatly expand our capabilities with dynamic parking pricing and develop demand forecasting methods.
Expected results and effects
Our hybrid algorithm was greatly improved by predicting traffic metrics extending over 2 hours in future time with fine granularity / high accuracy. Our MVP demonstrated considerable improvement to traffic flow on traffic light intersections. Our technology tools expanded to parking space demand and dynamic pricing. We attracted international interest for B2B applications. Our know-how expanded. We will have opportunity to file for additional IP.
Planned approach and implementation
We focused on the following objectives: 1)Defined and implemented a MVP 2)Searched/ collaborated with municipality for the MVP 3)Improved Machine Learning capabilities 4)Assembled an AWS secure MVP Back-end for an App or Web 5)Designed and outsourced an App/website Front-end 6)Executed MVP pilot (In Gothenburg) 7)Explored B2B opportunities with parking application (CES 2019, Barcelona, Paris) 8)Explored IP benchmarking, new IP opportunities, functionalities.