Open Light Solutions expands Renesas' Renesas autonomy platform, which will be on display at the Dibotics booth at CES 2018 (North Central Hall, C-5, French Motor Pavilion)
Today's optical processing, the need to combine high-performance processing platform and advanced embedded software.By combining Renesas high-performance image processing capabilities and low-power characteristics of automotive R-Car SoC, and Dibotics 3D SLAM (Simultaneous Localization And Mapping, simultaneous positioning and imaging) technology, the two companies jointly provide SLAM on Chip (Note 1) solution.
The SLAM on Chip enables 3D SLAM processing on a single SoC, a feature that previously had to be achieved only by a high-performance PC, which also enabled a 3D image to be achieved with just light data, With the elimination of inertial measurement units (IMUs) and global positioning system (GPS) data, this collaboration enables real-time 3D imaging systems with low power and high levels of functional safety in automotive systems.
As the automotive market prepares for the unmanned life, sensor technology is optimized for the features that unmanned vehicles demand, including real-time, high-precision sensing of the environment, precise vehicle positioning, and real-time sensor fusion. Still a major challenge.
Kodak has now become one of the key sensors in this field. Compared with other solutions such as cameras and radars, Kodak provides more accurate vehicle-surrounding obstacle sensing and a real-time electronic control unit (ECU) for vehicle control. Management. The rapid growth of the amount of data provided by the Shin-Kunda sensor technology has also increased the need for high-performance, real-time processing of these data.
Jean-Francois Chouteau, vice president of Renesas Electronics Global ADAS Center, said: "The seamless integration of Renesas Technology and Dibotics enables real-time, high-performance solutions for advanced photonics data processing with high-level FuSa capabilities Program. "" For drone, Renesas is working to optimize our end-to-end Renesas autonomy platform by working with innovative market partners. "Dibotics is providing Guitar to change the game's rules of thumb, We are pleased to announce the cooperation with them. "
Raul Bravo, CEO and co-founder of Dibotics, points out: "The light sensor will play a key role in all ADAS and autopilot functions, and will bring together the best of Renesas design expertise with the unique technology of Dibotics To provide the most cutting-edge performance and will allow OEMs, first-line suppliers and light manufacturers to reduce development costs and accelerate time-to-market, and we are excited to be working with market leaders such as Renesas. "
Unlike existing methods, the 3D SLAM technology implemented by Augmented LiDAR software from Dibotics requires only light to reach the 3D image from the data provided by the sensor. It does not require an IMU, GPS, or roulette encoder ), Eliminating the need for additional consolidation workloads, reducing BOM costs, and simplifying development efforts.In addition, the software also implements point-wise classification (Note 3), the shape and speed of moving objects , Track detection and chasing ?, and the integration of multiple sets of light (Note 4).
The R-Car SoC delivers Augmented LiDAR software from Dibotics through its high performance, low power consumption and high functional safety as set out in the FuSa standard of AS 26262. R-Car It is also part of Renesas' ground-breaking Renesas autonomy platform for ADAS and autopilot, providing a complete end-to-end solution from the cloud to sensing to vehicle control.
Dibotics will showcase its Augmented LiDAR solution at CES 2018 (Grand Place North Hall, CP-5, French Motor Pavilion) from January 9-12, 2018 in Las Vegas, Nevada.
(Note 1) SLAM on Chip (Simultaneous Localization and Mapping on Chip): SLAM is a computational algorithm that generates and updates maps of unfamiliar environments while chasing the vehicle's location. SLAM On Chip implements SLAM technology on SoC.
(Note 2) The wheel encoder measures the rotation of the wheel to get the speed of the car.
(Note 3) Automatic classification of each point provided by Guangda eliminates the need for machine learning, prior knowledge, or map data.
(Note 4) Real-time combination of multiple sets of optical data without calibration or synchronization.