Formula Technion

As part of the Formula Technion team, I contributed to the development of the Technion's first autonomous electric vehicle in 2023, focusing on autonomous systems. My main responsibility was implementing the FastSLAM algorithm in C++, enabling the car to simultaneously localize itself and map its environment using data from onboard sensors.

The car's perception system fused data from LiDAR and a camera. LiDAR provided precise distance measurements and 3D positional data, while the camera identified cone colors and shapes. Together, these inputs enabled the system to detect, classify, and localize cones, which defined the track layout. This real-time mapping was critical for path planning and decision-making.

To improve state estimation, we integrated Kalman filters into the SLAM process. These filters minimized noise in sensor readings and enhanced the car's ability to track its position accurately, even when facing rapid changes in movement or external disturbances.

One of the unique aspects of our design was the car's ability to adapt to its environment. During the boot process, the vehicle calculated key constants—such as friction coefficients—based on initial tests. This adaptation allowed the car to account for environmental factors like rain or surface irregularities, improving traction and handling without manual intervention.

To accelerate development and ensure reliability, we partnered with Cognata Ltd., a leading provider of autonomous vehicle simulation platforms. Their software enabled us to create a digital twin, a virtual replica of the car. The digital twin allowed us to test algorithms in a controlled environment before deploying them on the physical car. This approach reduced the risk of hardware failures and streamlined debugging by identifying issues early in the process.

The project was tested on a global stage during the 2023 Formula Student competitions in Europe, where teams from universities worldwide showcased their autonomous vehicles. The track was marked by cones, and the vehicle successfully navigated it autonomously, demonstrating real-time mapping, localization, and adaptability.

This project marked a significant milestone for the Technion, being the first year our team developed an autonomous vehicle. It combined advanced algorithms, sensor fusion, and adaptive control systems, all while balancing the constraints of a highly dynamic environment. For me, it was an opportunity to apply theoretical knowledge to a real-world challenge, work with a multidisciplinary team, and push the boundaries of what a student-led initiative can achieve.