A space dedicated to experimenting with innovative solutions based on AI algorithms and blockchain techniques to enable new models of mobility and smart cities.
,allowExpansion)
Best Practice
AI-enhanced Driver State Monitoring goes beyond sensors
Harnessing the potential of high-fidelity simulations, Reply delves into the future of autonomous driving solutions developing an innovative AI-enhanced and sensor-less approach for identifying impaired driving conditions.
#Impaired driving detection
#Driver state monitoring
#Artificial Intelligence on the edge
#Traffic scenario-based testing
Stay alert: Europe's 2026 drive with DSM
In order to reduce the number of road accidents caused by human factors, Driver State Monitoring (DSM) solutions have been developed. These systems are capable of detecting instances of impaired driving, such as drowsiness, distraction, driving under the influence, sudden illnesses, and aggressive driving. These solutions are also required by the European Regulation which, starting from 2026, it will be mandatory to equip all new vehicle types with the Advanced Driver Distraction Warning System (ADDWS).
,allowExpansion)
Reply's digital drive for enhanced monitoring
Reply has developed a driver state monitoring system based on the analysis of behavioural patterns (without the need for biometric data), and the creation of a Digital Twin of the driver. This model of nominal behaviour is adapted to each specific driver's driving style by collecting telemetric data on an AWS platform and analysing the historical usage on the road.
To validate this algorithm, a testing framework has been developed, which includes an in-cabin monitoring system capable of analysing drivers' biometric data and validating instances of inappropriate driving activities.
Innovative framework features
Privacy-centric
Detects impaired driving without the necessity to tap into biometric information.
Driver personalization
Create a Digital Twin tailored to each verified driver, reflecting their unique driving patterns.
On-device processing
Engineered to operate directly on the vehicle's ECU, minimising lag and bypassing network limitations.
Self-regulating labelling
Streamlines the data tagging and system verification process by using real-time insights gathered by an in-cabin camera monitoring setup.
Seamless vehicle connectivity
Anchored on a unified vehicle connectivity framework, it's adept at transmitting telemetry insights to distant databases and facilitating over-the-air model updates.
Advanced biometrics
Utilises computer vision techniques to derive biometric insights, ensuring precise assessment of driver distraction levels.
X-in-the-loop integration
A synergistic, scalable co-simulation environment combined with a realistic vehicle model is the basis for a versatile tool that can be used in development and validation, especially for assessing traffic scenarios.
Synthetic data generation
Generation of high-fidelity vehicle and traffic data from a simulation for automotive system development and validation.
Cost-Efficiency &
Time-Saving
Unlike real-world data acquisition which involves extensive sensorisation and experimentation, simulated environments can produce massive amounts of data in a fraction of the time and cost.
)
Safety
.
Simulated driving scenarios mitigate the risks tied to real-world anomalies, ensuring driver safety isn't compromised during data collection.
Reproducibility &
Scalability
With simulations, one can reproduce diverse driving scenarios, ensuring a holistic dataset that mirrors various real-world conditions.
)
Dynamic Interaction
Emulation
From capturing subjective behaviour with a real driver in the loop, to interactions with smart road infrastructure, simulations offer a detailed and realistic portrayal of on-road conditions.
Legislation
Compliance
The certification processes of these simulations are already aligned with EU GSR legislation, paving the way for more streamlined, legal, and standardised data collection processes.
AREA42 LAB
From ideas to innovative projects
The autonomous driving technology and high-fidelity simulation are tangible realities at Area42 labs: the Reply’s development centre is dedicated to experimenting with and transforming creative ideas into reality and it is where real prototypes can be felt, tested, and adapted to any business context.
This is the future of autonomous driving, and it's being shaped right now at Area42.
You may also be interested in
,allowExpansion;Resize,width=660)
,allowExpansion;Resize,width=660)
;Resize,width=660)
,allowExpansion;Resize,width=660)
Concept Reply, part of the Reply Network, is a leader in IoT solutions. They support and advise customers across Automotive, Manufacturing, Smart Infrastructure and beyond in all aspects of Internet of Things (IoT) and Cloud Computing. From designing and developing customized IoT solutions to implementing and managing them seamlessly, Concept Reply helps customers unlock the potential of IoT.
Sensor Reply
Sensor Reply is a company of Reply Group focused on AI-powered IoT applications. The company’s mission is to supply customers with data-driven software applications and engineering services based on the integration of IoT edge platforms and Artificial Intelligence methods. The company integrates the knowledge of forefront methods in Machine Learning and Deep Learning, with a deep experience in system control, diagnostic, modelling and edge computing.
Storm Reply is an AWS Premier Consulting Partner and the specialist for professional cloud computing services in the Reply network. Through consolidated expertise and many years of experience in the creation and management of Data and Analytics solutions, Infrastructure as a Service (IaaS), Software as a Service (SaaS), and Platform as a Service (PaaS), Storm Reply supports global clients in the implementation of cloud-based systems and applications. Our goal is to use the cloud to create added value for mid-sized and large companies and to help them exploit the full potential of the cloud.