After 2-year-long scientific research, TCR App Mobility has developed a mobile application which can achieve empirically-proven 94% effectiveness in prediction of road traffic accidents caused by tiredness and inattention. Outside of a vehicle, it employs sophisticated algorithms based on machine learning in order to describe the user’s distinctive brain activity (by analysing his/her lifestyle and health conditions). This process enables us to figure out the individual susceptibility to abrupt attention falls of a given driver.

 

After entering a vehicle, TCR App measures almost a hundred parameters designed to represent the user’s subconscious behavioural patterns (mainly associated with his/her trajectory, acceleration and biometric traits). We have implemented an advanced recurrent deep neural network (LSTM RNN) which allows us to translate collected data (e.g. regarding how the driver's trajectory or acceleration fluctuates in different travel circumstances) into probabilistic mathematical models - subsequently utilized to diagnose in what way the user's organism behaves a few minutes before an approaching attention fall. Hence, we can forecast the future collisions thanks to the observation of previously-discovered tendencies.

Finally, our AI-driven decision-making algorithm empowers us to set a specific moment in time when TCR App’s sound alert will be emitted. Whenever a dangerous tendency in registered data is identified, our mobile application is poised to directly impact the user’s brain via voice alarms devised to interrupt the natural process of losing attention prior to any of its negative consequences (hence ensuring the sufficient level of consciousness through the entire journey). TCR App attains research-proven 94% effectiveness of its predictions after just 3 months of contact with a particular customer.