Science and Validation
Legion simulations are anchored by the largest and most complete pedestrian measurement programme ever undertaken.
Since 1999, research and development carried out by Legion and our research institute (Maia) has gathered more than 1000 hours of video footage of pedestrians in 14 cities on three continents, covering more than 40 discrete contexts.
No other simulation company has this volume of real-world data and Legion uses it to develop accurate computational algorithms, which are calibrated according to our observations and validated against genuine pedestrian movements.
Thanks to this programme of measurements, we don't need to construct unsubstantiated analogies with computational fluid dynamics, gravity, ant colonies or electromagnetic theory. Legion’s scientific model has intuitive appeal: our pedestrians move on the principle of ‘least effort’. Each chooses its next step in an effort to find the best compromise between directness of path, speed and comfort.
These decisions take into account an agent’s preferences and objectives as well as the context, environment and other pedestrians around them. The result is a solidly justifiable model that generates simulations with predictive power in a very wide range of situations.
Not content to rest with the observations and results gathered so far, we are continuing with our measurement programme, the scientific analysis of our observations, and we continue to contribute regularly to current academic debates.
Since 1999, research and development carried out by Legion and our research institute (Maia) has gathered more than 1000 hours of video footage of pedestrians in 14 cities on three continents, covering more than 40 discrete contexts.
No other simulation company has this volume of real-world data and Legion uses it to develop accurate computational algorithms, which are calibrated according to our observations and validated against genuine pedestrian movements.
Thanks to this programme of measurements, we don't need to construct unsubstantiated analogies with computational fluid dynamics, gravity, ant colonies or electromagnetic theory. Legion’s scientific model has intuitive appeal: our pedestrians move on the principle of ‘least effort’. Each chooses its next step in an effort to find the best compromise between directness of path, speed and comfort.
These decisions take into account an agent’s preferences and objectives as well as the context, environment and other pedestrians around them. The result is a solidly justifiable model that generates simulations with predictive power in a very wide range of situations.
Not content to rest with the observations and results gathered so far, we are continuing with our measurement programme, the scientific analysis of our observations, and we continue to contribute regularly to current academic debates.