The First Full-Stack Implementation of JPEG XS RAW
intoPIX, a leading provider of low-latency image and video compression technologies, today announced that dSPACE will showcase a high-performance data logging solution integrating TicoRAW, the most advanced implementation of JPEG XS RAW ISO standard, at AutoSens US.
As the automotive industry accelerates the development of advanced driver assistance systems (ADAS) and autonomous driving (AD), the need to efficiently capture and manage massive volumes of sensor data has become critical. Modern test vehicles generate tremendous amounts of RAW data from high-resolution cameras, lidar, and radar sensors, creating major challenges in storage, bandwidth, and data processing.
dSPACE’s data logging platforms are specifically designed to address these demanding requirements. Their modular in-vehicle systems can record synchronized multi-sensor data streams with very high bandwidth—more than 50 Gbit/s—and support raw sensor interfaces and hardware acceleration such as GPUs and FPGAs, enabling cutting-edge data acquisition for autonomous driving validation and development.
Bringing RAW Compression to Automotive Data Logging
Through the integration of TicoRAW into the data logging pipeline, dSPACE demonstrates how RAW sensor data can be efficiently compressed at the source while preserving full image fidelity.
intoPIX TicoRAW is JPEG XS RAW, the only ISO standardized format for RAW image data compression.
“The integration of JPEG XS RAW by dSPACE is a key milestone in the deployment of compression across the automotive market.” Said Pascal Pellegrin, CTO of the intoPIX Automotive Group. “The challenges raised by the massive amounts of data generated to develop and validate ADAS functionalities require innovative solutions. With JPEG XS RAW, dSPACE and intoPIX are enabling the adoption of efficient storage and transmission at scale.”
Key benefits of the integration include:
- Mathematically lossless or visually lossless compression of RAW Bayer data
- Significant reduction in storage requirements onboard the vehicle
- Extended recording time during test drives
- Faster data transfer to data centers and cloud infrastructure
- Maintained data quality for AI training and validation
Proven Integration on dSPACE Data Logger Platforms
Initial work between intoPIX and dSPACE has already demonstrated successful integration of TicoRAW encoding within the data logger environment.
Early results show efficient RAW data compression conveniently integrated into the RTMaps software, including CPU-based and GPU-based implementations, opening the door to handling multiple high-resolution camera streams simultaneously in real-time. This confirms that TicoRAW can be seamlessly integrated into advanced data logging architectures without disrupting existing workflows.
Accelerating Data-Driven Development for ADAS and Autonomous Driving
dSPACE solutions are widely used by automotive OEMs and Tier‑1 suppliers to capture real-world driving data for validation, scenario generation, and AI training. Their data logging systems enable synchronized acquisition of large volumes of sensor data during test drives—an essential requirement for developing robust perception and sensor fusion algorithms.
By combining this capability with TicoRAW, engineers can:
- Capture more data per vehicle and per test campaign
- Reduce infrastructure costs for storage and cloud transfer
- Ensure access to high-quality RAW datasets required for AI model training
- Improve the overall efficiency of the ADAS/AD development pipeline
“At dSPACE, we excel in delivering high-performance data acquisition systems for next-generation vehicle development,” said Jacob Perrin, Manager in ADSE at dSPACE “Integrating RAW compression technologies like TicoRAW further enhances the scalability and efficiency of our data logging solutions.”
Live Demonstration at AutoSens US
intoPIX (booth #313) will showcase this integrated data logging solution at AutoSens US, where visitors can experience how TicoRAW enable efficient, scalable RAW data logging for next-generation automotive systems.
This demonstration highlights a key step toward standardized, efficient RAW data handling across the automotive ecosystem.