The efficiency and execution time of a road scene evaluation subsystem directly influences a self-driving car's control effectiveness. This article presents a novel approach to fuse data from various sensors (camera, LIDAR, radar, IMU) for pedestrian and vehicle tracking. Our approach, thanks to modern methods of image processing and power of GPU for LIDAR processing, achieves 25Hz update frequency for tracking. The system has been tested on the DiDi-Udacity Self-Driving Challenge datasets.

Car racing control is often assumed to take place in isolation, although the greatest strategic challenges arise in the presence of other cars. Our solution uses a combined nonlinear system model which includes the four-wheeled vehicle dynamics model and point-mass model of its opponent car. This proposed combined model allows the optimal control scheme to automatically find an overtaking strategy. The FORCES code generation tool to handle the consequent computational complexity of the Nonlinear Model Predictive Controller (NMPC). We demonstrate the practicality of the suggested method by performing real-time control in a racing simulation software.

In the near future anthropomorphic robots will turn into an important part of our everyday routine. To successfully perform various tasks these robots require stable walking control algorithms, which could guarantee dynamic balance of the biped robot locomotion. Our research is focused on the development of locomotion algorithms which could provide effective anthropomorphic walking of a robot. As a target robotic platform we utilize an experimental model of a human-size robot - a novel Russian robot AR-601M. In this paper we introduce AR-601M robot and present a model of a biped robot with 11 DoF which simulates a simplified AR-601M robot. The simulation model is implemented in Matlab/Simulink environment and uses walking primitives in order to provide a dynamically stable locomotion.