Smart Eye Unveils Multimodal In-Cabin AI Agent Demo

On September 23, Smart Eye announced a technology demonstration centered on a multimodal in-cabin AI agent designed to understand context, determine the appropriate next action, and respond accordingly. The technology will make its public debut at InCabin Europe 2026, giving the company a platform to demonstrate how multiple sources of in-vehicle information can be combined to support more responsive automotive AI. The demonstration focuses on interpreting driver and cabin conditions rather than relying on a single input, bringing together information that can help the system determine what is happening inside the vehicle and how it should react.

Multimodal Inputs Support Context-Aware Responses

The AI agent uses inputs including voice, driver identity, gaze, attention, emotions, driver condition, and occupancy, together with information collected by the vehicle’s interior and exterior sensors. Combining these inputs gives the system a broader view of the vehicle environment and the driver’s state. According to Smart Eye, this approach can reduce the computing power required while enabling the automotive AI agent to respond more quickly and effectively. The technology therefore links multiple contextual signals instead of treating individual inputs as isolated sources of information within the vehicle.

Integration With ADAS and Vehicle Systems

The platform combines driver behavior data with information from ADAS, vehicle systems, and other sensors, creating a broader information layer for the in-cabin AI agent. This integration allows the system to consider driver-related information alongside data generated by other vehicle functions and sensing systems. The approach is designed to support contextual decision-making, where the agent can use available information to determine an appropriate response. The demonstration highlights how in-cabin intelligence can interact with wider vehicle data rather than operating independently from ADAS and other vehicle systems.

Flexible AI Model and Voice System Support

The platform is designed to work with different AI models, voice systems, and agents operating both inside the vehicle and in the cloud. This flexibility allows the technology to combine multiple AI and voice capabilities rather than being limited to one model or system. By supporting both in-vehicle and cloud-based agents, the platform can connect different sources of computational and AI functionality. The architecture also reflects the broader role of software-based intelligence in vehicles, where information from sensors, vehicle systems, and driver monitoring technologies can be brought together to support more context-aware interactions.

InCabin Europe 2026 Demonstration

The technology will be publicly presented for the first time at InCabin Europe 2026. The event will provide the setting for Europe-based industry audiences to see how the multimodal in-cabin AI agent combines driver information, vehicle data, and sensor inputs. The demonstration is focused on the technology’s ability to understand context, identify the next action, and respond accordingly. Its public presentation also provides an opportunity to show how different AI models, voice systems, and agents can operate together across in-vehicle and cloud environments while using a range of available vehicle inputs.

Industry Impact & Outlook

The demonstration reflects the automotive industry’s continuing development of AI systems that combine driver monitoring, cabin sensing, ADAS information, and vehicle data into more context-aware software experiences. For automakers and technology suppliers, the ability to connect different AI models, voice systems, and agents across vehicle and cloud environments could support more flexible approaches to in-cabin intelligence. The emphasis on reducing computing requirements while improving response speed is also relevant to practical deployment. The next important step will be demonstrating how the technology performs across real vehicle scenarios and how its multimodal capabilities can be integrated with broader vehicle systems.

Frequently Asked Questions

What is Smart Eye’s new in-cabin AI technology?
Smart Eye’s new technology is a multimodal in-cabin AI agent designed to understand vehicle and driver context, determine the next action, and respond accordingly using multiple data sources. It combines voice, driver identity, gaze, attention, emotions, driver condition, occupancy, and information from interior and exterior sensors. The platform also integrates driver behavior data with ADAS, vehicle systems, and other sensors. It can operate with different AI models, voice systems, and agents in the vehicle and in the cloud, while the company says the approach can reduce computing requirements and improve response speed.