MEDIATOR · Project

Smart System That Decides When the Driver or the Car Should Be in Control

transportPilotedTRL 6

Imagine you're driving a car that can handle some tasks on its own — like cruise control on steroids — but it can't handle everything yet. The tricky part is knowing when YOU should take over and when the car is doing fine. MEDIATOR built a smart referee that sits between you and the car's autopilot, constantly watching how alert you are, how capable the automation is, and what's happening on the road. When either you or the car starts struggling, it steps in and hands control to whoever can handle the situation better.

By the numbers

consortium partners across 6 countries

industrial partners in the consortium

56%

industry participation ratio

EUR 6,461,615

total EU contribution

project deliverables produced

demo-level deliverables (in-vehicle + lab prototypes)

countries represented in consortium

The business problem

What needed solving

Vehicles with partial automation (Level 2-3) create a dangerous gap: the driver is supposed to monitor the system and take over when needed, but attention drifts, fatigue sets in, and people either trust the automation too much or too little. This causes accidents that neither full manual driving nor full automation would produce. Car makers, fleet operators, and insurers all need a reliable way to manage who is in control — and switch safely — during this transition period that will last years.

The solution

What was built

MEDIATOR produced final in-vehicle prototypes and a final integrated lab prototype — an AI-based system that monitors driver state, automation capability, and road context in real time to manage control transitions. Across 24 deliverables, the project developed the mediating intelligence, driver monitoring algorithms, and evaluation tools, all tested through simulation, driving simulators, and real on-road studies.

Audience

Who needs this

Automotive OEMs developing Level 2-4 automated driving featuresTier-1 ADAS suppliers building driver monitoring and handover systemsCommercial fleet operators managing semi-automated truck or bus fleetsMotor insurance companies pricing risk for partially automated vehiclesTransport safety regulators developing standards for human-automation interaction

Business applications

Who can put this to work

Automotive OEMs and Tier-1 Suppliers

enterprise

Target: Vehicle manufacturers and ADAS component suppliers developing Level 2-4 automated driving systems

If you are an automotive OEM or Tier-1 supplier struggling with safe handover between driver and automation — this project developed in-vehicle prototypes with real-time AI that monitors driver alertness, automation capability, and road context simultaneously. With 9 industrial partners already involved in development across 6 countries, the technology was validated through on-road studies in actual vehicles. This could reduce your liability exposure from automation-related incidents during the critical transition period to full self-driving.

Commercial Fleet Management

mid-size

Target: Trucking and logistics companies operating semi-automated vehicle fleets

If you are a fleet operator worried about driver fatigue and distraction in trucks with advanced driver assistance — this project built a system that personalises its monitoring to each driver's competence level and detects degraded performance in real time. The 16-partner consortium tested prototypes in actual vehicles, not just simulators. For fleets running long-haul routes, this means fewer accidents caused by the grey zone where neither driver nor automation is fully in charge.

Automotive Insurance

enterprise

Target: Motor insurers developing usage-based and ADAS-aware policies

If you are an insurer trying to price risk for vehicles with partial automation — this project created tools that continuously evaluate driver state, automation status, and driving context. The system generates real-time data on who was effectively in control at any moment, which is exactly the missing piece for determining liability after an incident. With EUR 6,461,615 in EU funding and 24 deliverables produced, the research base is substantial for building actuarial models around human-automation interaction.

Frequently asked

Quick answers

What would it cost to license or integrate this technology?

The project did not publish specific licensing terms. With 16 consortium partners — including 9 from industry and 2 SMEs — licensing arrangements would need to be negotiated with the coordinator SWOV in the Netherlands and relevant IP-holding partners. Given the EUR 6,461,615 EU investment, expect serious commercial terms.

Can this scale to production vehicles, or is it still lab-only?

MEDIATOR went beyond the lab. The project produced both a final integrated lab prototype and final in-vehicle prototypes, tested through simulation, simulator, and on-road studies. This puts it closer to production readiness than most research projects, though automotive certification and series production integration would still be needed.

Who owns the intellectual property and how can I access it?

IP from EU-funded RIA projects is typically owned by the consortium partners who generated it. With 9 industrial partners already in the consortium, some IP may already be heading toward commercial products. Contact the coordinator SWOV (Netherlands) for licensing discussions.

Does this comply with current vehicle safety regulations?

MEDIATOR was designed with market exploitation in mind and actively involved the automotive industry during development. However, the project ended in April 2023, so alignment with the latest EU General Safety Regulation and UN/ECE automated driving regulations would need to be verified for current compliance.

How long would integration into an existing ADAS platform take?

The project ran for 4 years (2019-2023) from concept to in-vehicle prototypes. For a company with an existing ADAS platform, integration of the mediating logic and driver monitoring AI would likely be significantly shorter, but no specific integration timeline was published in the project data.

Does it work with existing vehicle sensor hardware?

Based on available project data, MEDIATOR used state-of-the-art AI to evaluate driver state, automation status, and driving context in real time. The system was tested in actual vehicles, suggesting compatibility with existing automotive sensor platforms, though specific hardware requirements would depend on the vehicle platform.

Consortium

Who built it

This is a strong, industry-heavy consortium with 16 partners across 6 European countries (Germany, Israel, Italy, Netherlands, Norway, Sweden). The 56% industry ratio — 9 out of 16 partners — signals that this was built with commercial application in mind, not just academic publication. The mix of 3 universities and 4 research institutes provided the scientific backbone, while the industrial majority ensured the prototypes were grounded in real automotive requirements. The coordinator, SWOV (Netherlands), is a well-known road safety research institute classified as an SME. Having partners in key automotive markets like Germany, Sweden, and Italy strengthens the path to market adoption.

STICHTING WETENSCHAPPELIJK ONDERZOEK VERKEERSVEILIGHEID SWOVCoordinator · NL
AUTOLIV DEVELOPMENT ABparticipant · SE
STICHTING KONINKLIJK NEDERLANDS LUCHT - EN RUIMTEVAARTCENTRUMparticipant · NL
CENTRO RICERCHE FIAT SCPAthirdparty · IT
KONGSBERG MARITIME ASparticipant · NO
ZENSEACT ABparticipant · SE
BEN-GURION UNIVERSITY OF THE NEGEVparticipant · IL
MAGNA ELECTRONICS SWEDEN ABparticipant · SE
STATENS VAG- OCH TRANSPORTFORSKNINGSINSTITUTparticipant · SE
TECHNISCHE UNIVERSITAET CHEMNITZparticipant · DE
STELLANTIS EUROPE SPAparticipant · IT
CYGNIFY BVparticipant · NL
KONGSBERG NORCONTROL ASthirdparty · NO
TECHNISCHE UNIVERSITEIT DELFTparticipant · NL

How to reach the team

SWOV — Netherlands road safety research institute. Use the CORDIS contact form or search for MEDIATOR project coordinator at SWOV for direct contact.

Order a report on this consortium See STICHTING WETENSCHAPPELIJK ONDERZOEK VERKEERSVEILIGHEID SWOV profile →

Next steps

Talk to the team behind this work.

Want an introduction to the MEDIATOR consortium? SciTransfer can connect you with the right partner for your specific use case — whether that's the driver monitoring AI, the mediation logic, or the in-vehicle integration expertise.

Order a report on this topic More in Transport & Mobility