Advanced algorithms to support high-precision LiDAR

The success of self-driving cars is also facing many challenges. Among them, real-time high-definition sensing of the surrounding environment, accurate knowledge of vehicle location, and sensor fusion in real time are the three major challenges currently facing. Among the many key sensors that realize the perception of advanced 3D environments, LiDAR (Lidar) sensors are undoubtedly becoming more and more important. In order to solve these problems, Dibotics and Intempora have announced the establishment of a technology partnership. Dibotics' advanced real-time LiDAR data processing algorithms and precise SLAM (instant positioning and map construction) technology will be embedded in Intempora. The company's RTMaps platform provides a common solution for engineers who are facing these challenges. Dibotics specializes in innovative real-time LiDAR processing software. Dibotics has developed an embedded LiDAR software called "Augmented LiDAR" that provides real-time, advanced LiDAR data processing.

Augmented LiDAR algorithm map and LiDAR raw data graph comparison Augmented LiDAR software solves the environmental perception challenges faced by mobile robots, especially autonomous vehicles, mainly through precise positioning/self-driving, 3D mapping, obstacle detection and tracking, Sensor fusion without calibration and point intelligent classification (passages, road signs, lights, moving objects, etc.).

Video: Augmented LiDAR algorithm map and LiDAR raw data graph comparison Unlike traditional technology solutions, Augmented LiDAR software only requires data directly from LiDAR sensors, without the need for additional inertia measurement unit (IMU), GPS or wheel speed sensors. data. All calculations used in the software are deterministic and do not require artificial intelligence or machine learning.

Intempora's RTMaps platform is an asynchronous high-performance platform designed for the current multi-sensor fusion, enabling engineers and researchers to take advantage of efficient and easy-to-use architecture to quickly and robustly develop products and promote advanced driving. Design and Application of Auxiliary Systems (ADAS) and Highly Automated Driving (HAD) "RTMaps can help customers effectively utilize our 3D SLAM and classification recognition software in seconds to seamlessly interface with many different sensors, software and hardware platforms. The first customers were highly appreciative of the platform after the application,” commented Raul Bravo, co-founder and CEO of Dibotics. "The addition of the Augmented LiDAR algorithm to the RTMaps platform allows us to further meet the ever-increasing demands of our customers and partners. As a ready-made RTMaps building block, this 3D SLAM algorithm can help solve key challenges such as context awareness and real-time target classification. The cooperation of Dibotics allows us to continue to provide the best autonomous driving technology for the market,” said Intempora CEO Nicolas du Lac. Dibotics and Intempora agree that the autopilot industry needs new perspectives and tools to help develop safer, better quality solutions. Dibotics' embedded modules in the RTMaps platform will help engineers working on autonomous driving and ADAS applications accelerate their time-to-market and streamline the development of specific applications. About Dibotics

Based in Paris, France, Dibotics is committed to providing innovative real-time LiDAR data processing. Dibotics was founded in 2015 by Raul Bravo and Olivier Garcia, and two consecutive entrepreneurs have 15 years of experience in LiDAR data processing for mobile robot applications. Dibotics has an original technical solution that is different from the traditional model. It does not pick sensors (supporting 2D & 3D LiDAR, 2D & 3D sonar/radar, 3D camera) 3D SLAM technology and Augmented LiDAR algorithm, creating the first set of data obtained by relying solely on the sensor itself, enabling point intelligence classification, Target detection/tracking and solutions that eliminate the need for advanced features such as sensor fusion. The implementation of all these advanced features does not require any machine learning and can be embedded in small-sized, low-power integrated circuits (SLAM on Chip, on-chip SLAM system).

According to a report previously reported by Memes Consulting, in July 2016, Dibotics and LiDAR leader Velodyne LiDAR signed a cooperation agreement. The two companies will jointly develop LiDAR sensors in the fields of UAV mapping, driverless cars and mobile robots. application. Dibotics will provide consulting services to Velodyne LiDAR's customers who require 3D SLAM software and Augmented LiDAR algorithms.

About Intempora

Intempora is a software company founded in 2000 based on the research and development work of the Mines-ParisTech Robotics Research Center since 1998. Intempora has more than 15 years of experience in advanced software solutions for the automotive industry, especially in real-time multi-sensor fusion applications. Multi-sensor fusion technology plays a very important role in robotics, ADAS and highly automated driving applications.