ICASIS: A Scalable Simulation Framework for Validating In-Cabin Monitoring Algorithms

This session presents the evolution of ICASIS from a desktop prototype to a cloud-native SaaS platform for simulating and validating in-cabin monitoring systems. We detail how ICASIS enables OEMs and Tier 1 suppliers to generate synthetic datasets across diverse vehicle configurations, occupant models, and camera sensor types—including RGB and IR modalities. The talk will focus on the technical challenges of modeling IR sensor behavior, bridging the domain gap between synthetic and real-world data, and aligning simulation outputs with industry validation standards. We’ll also discuss our architectural transition to a scalable SaaS infrastructure, supporting modular sensor pipelines and customer-specific simulation workflows for robust algorithm training and benchmarking.