Our Engine Block

The Mantis-Module, is structured around our patented ALGORITHM.From its inception, our Mantis-Module was built on solid mathematical principles. From early on, it demonstrated its high accuracy and its exceptional resilience handling increased traffic flows. It is a totally different innovative approach to the traffic flow problems braking the industry's status quo combining AI, Machine Learning and Stochastic modelling.

As a matter of fact, all current traffic models sooner or later exhibit a compounding inability to respond to excessive gridlock conditions. The larger the actual vehicle densities become, the larger the generated error they produce and the more unreliable solutions they offer.

What Ximantis brings to the table and makes it such a unique proposition is the demonstrated fact that while, with increased density, our competitor models deviate, the Ximantis solution shows complete agreement to the exact known solution. Our Stochastic approach alone is the only known case where for high densities, as navigation options demand peaks, an exact extraction solution is finally possible.

The road to ximantis

Find More About Us

Turning the Gears

key elements

Minimum Viable Product

Our Gothenburg, Sweden MVP trials with traffic lights and city road density variations, demonstrated 99% prediction accuracy in 30 minute forecasting intervals and repeated similar high marks for One and Two hours test intervals.

Data Sources

We collect  data (location, speed, direction, date-time) from GPS devises either onboard vehicles or smart devices, Traffic Video Cameras, RFID readers and many other sources and supplement it with additional data from traffic control centers, live shared data among users, weather forecasting and special scheduled events.

Look Ahead Value

Ximantis provides more than 30 minutes advance warning of adverse traffic conditions ahead relying on real time traffic data from GPS and other sources, weather, traffic accident reports and other connected user network input to provides very reliable and accurate Travel Times and ETAs.

Cloud Infrastructure

Operational Setup and Management through our White Label Cloud Platform collecting, processing  and disseminating information from and to individual Users, Fleets and Traffic Control.

High Density Capacity

Our hybrid algorithm has an uncanny ability to handle overloaded traffic conditions contrary to all the other competing models. It maintains its accuracy for best optimized use of available road-space and exit routes even in increased high density traffic situations.

Proof of Concept

Concept validation with Field trials on US highway 101, Interstate 880 USA and Gothenburg, Sweden. Full presentation at ISTTT20 conference in 2013.

Proven Forecasting Concept

Our Machine Learning / Stochastic hybrid model has demonstrated its ability for accurate forecasting from only real time traffic data alone. Historical data is used for initial training. Afterwords, it relies on real time traffic data, captured live and supplemented with weather data, live traffic accident reports, social networking data shared among users, etc.