Technology
We offer a fully autonomous medical diagnostic system (agentic AI) delivered to users via a mobile application. It can predict and prevent road traffic accidents caused by tiredness and inattention (achieving research-proven 94% effectiveness within just 3 months of contact with a given individual). Our technology also enables presymptomatic medical diagnostics of prevalent neurodegenerative diseases (e.g., Parkinson's disease approximately 5 years before the first tremor) on the basis of subtle driving anomalies.
SUMMARY
Fundamental Neuroscience
Two years of scientific research (including the foundational model training workflow) with novel findings about human cognition. This process combined mathematical discoveries, VR simulations, real-world testing and neuroscience-related experiments — allowing us to introduce an original mathematical diagnostic paradigm characterized by unprecedented sensitivity and effectiveness
Mathematical Ground Truth
One of the largest proprietary datasets of digital biomarkers (focused on attention stability and high-dimensional health phenotyping), directly correlated with the way the human brain allocates its resources at the most fundamental level
Technological Engine
Proprietary ML models (a purpose-built LSTM and a Temporal Geometry-Aware Transformer in two distinct computational paradigms), an original adversarial inverse reinforcement learning framework, and a diagnostic-grade training workflow that effectively combined localized brain activity (EEG, fNIRS), biometric signals (sEMG, GSR, HRV, eye tracking), and operational data streams (CAN, CoP, IMU) with micro-adjustments of driving modalities and in-vehicle behavioural patterns
Proprietary Signal Processing
Feature engineering and data collection methods designed to successfully separate relevant signals (related to the user's health phenotypes and digital biomarkers) from high-noise environmental artefacts — optimized for older mobile devices and inherently enabled to suppress external factors, such as road irregularities, weather, worn vehicle components (e.g., brake discs, tires), driver respiration, arterial blood flow, and even random events like in-seat movements of a passenger
Edge Personalization
On-device individual cognitive decoder implemented as a static behavioural adapter (a compact set of algorithms tailored to the user's brain, requiring no local ML computing) and regularly actualized after data processing in the cloud
Custom Data Flywheel
Continuous evolution of the model thanks to raw, anonymized data from multiple users of our mobile application — aggregated in the cloud and utilized either to update the foundational architecture or to create its specialized variant
Seamless User Journey
No action is required from the user beyond installing our fully autonomous mobile application that operates in the background (regardless of whether the phone is mounted on the dashboard, placed anywhere within the vehicle, or kept in a pocket) and, whenever necessary, [1] emits sound alerts designed to avert attention lapses at least three minutes before any discernible cognitive impairment, or [2] provides a comprehensive medical assessment or a conclusive diagnosis of neurodegenerative disorders (PD, AD) 4–8 years before the onset of clinical symptoms
INTRODUCTION
TCR App Mobility offers a fully autonomous medical diagnostic system (agentic AI) delivered to users via a mobile application. It can predict and prevent road traffic accidents caused by tiredness and inattention (achieving research-proven 94% effectiveness within just 3 months of contact with a given individual). This process is made possible by converging raw IMU and GPS measurements (acquired both inside and outside the vehicle through built-in smartphone sensors) with infinite-dimensional modelling – leveraging our proprietary, empirically-derived mathematical constructs that are entirely independent of driving experience, advanced driver assistance systems (ADAS), mechanical issues (e.g., worn tires), environmental conditions (e.g., heavy rain, road irregularities), random events (e.g., in-seat movements of a passenger) and temporary or permanent vehicle changes.
Our solution emits sound alerts whenever a dangerous data pattern is detected — averting the natural lapses in attention at least three minutes before any discernible cognitive impairment. It ensures sustained awareness throughout the entire journey.
Once the collected raw data (IMU and GPS measurements) is cleaned and enriched via advanced feature engineering, it is processed through our proprietary "analytical neural network" architecture (a purpose-built LSTM with a tailored attention mechanism and a domain-specific cost function). Moving beyond standard supervised learning, we employ our novel adversarial inverse reinforcement learning (AIRL) framework — enhanced by latent behavioural embedding capabilities — to automatically infer the individual driving reward function of each user.
Central to our predictive methodology is an original mathematical construct established through two years of foundational scientific research — referred to as the "diagnostic function". This non-linear operator explicitly allows us to determine when the user's driving patterns are impacted by impending attention lapses by mapping AIRL-derived driving reward functions into a specialized, infinite-dimensional feature space associated with a positive-definite kernel (the "diagnostic space").
By treating driving as an optimization problem, we can mathematically model the underlying logic behind the user's subconscious decisions. Even a marginal shift in the user's neurological state triggers a non-linear jump in the driving reward function's raw values — analogous to how a single character change completely alters a cryptographic hash. By transforming these constructs into the infinite-dimensional "driving phenotypes" within our diagnostic space, we achieve topological continuity: a minor change in the user's physiological condition manifests as a proportionally small evolutionary pattern, enabling reliable tracking of abrupt attention lapses.
The predictive power of our system lies in its ability to monitor the evolution of driving phenotypes over time. By analysing the short-term drifts of their distribution and the fluctuations of their internal structure within our diagnostic space, we can identify the transient cognitive transition within the brain.
We have built one of the largest datasets of digital biomarkers focused on attention stability. Our system utilizes a proprietary Temporal Geometry-Aware Transformer (referred to as the "diagnostic neural network") trained on synchronized EEG, fNIRS, eye tracking (saccadic latency, fixation precision, microsaccade dynamics), sEMG, GSR, HRV, pressure mapping, and telematics data acquired within our custom-engineered VR driving simulator. Therefore, the temporary alteration of an individual driving phenotype serves as a unique representation of precisely defined neurophysiological processes (e.g., those detected during our ML training within the fronto-parietal attention networks, the default mode network, or the locus coeruleus-norepinephrine system).
Moreover, our algorithms are able to identify subtle changes in driving habits that are characteristic of prevalent neurodegenerative diseases (PD, AD) approximately 5 years before the onset of clinical symptoms (allowing the development and early start of disease-modifying therapies).
COMPETITION AND USP
The vast majority of the population is not protected in any systematic way against tiredness and inattention while driving. Our indirect competitors primarily include the vehicle-embedded safety systems (developed by car manufacturers), which utilize eye and lane-tracking cameras to detect excessive tiredness once it has already begun (without any real predictive abilities). These systems can also be acquired independently as either a device or a mobile application. In terms of medical diagnostics, our framework establishes an entirely new diagnostic paradigm.
Our technology:
[1] is based on science-backed custom deep learning models that are kept as trade secrets (allowing 94% effectiveness in prediction of road traffic accidents caused by inattention and guaranteeing a 5-year PD diagnostic lead time needed to deter future market entrants);
[2] can automatically describe the internal energy and attention allocation within the human brain, correlating non-invasive functional neuroimaging and built-in smartphone sensors;
[3] embeds a clinical-grade tool in a consumer-grade device by mapping complex neurophysiological impairments (e.g., prodromal stages of PD) into longitudinal evolutionary patterns of AIRL-derived driving reward functions in our infinite-dimensional feature space — our original discovery that allows us to both identify the functional impact of the first 5–10% of neuronal degeneration and predict approaching attention lapses;
[4] is implemented as a fully autonomous mobile application (agentic AI) with a proprietary data flywheel and a unique multi-modal dataset of digital biomarkers;
[5] operates in the background of the user's daily routine and transforms regular driving into a part of the medical diagnostic system, ensuring long-term compliance without requiring any lifestyle changes or conscious interaction from the user (eliminating the engagement decline and individual biases);
[6] introduces an original mathematical diagnostic paradigm characterized by unprecedented sensitivity and effectiveness, directly correlated with the way the human brain allocates its resources at the most fundamental level.