Maria Rosario Garcia et al., International Journal of Advanced Trends in Computer Science and Engineering, 11(3), May – June 2022, 121 – 126 121 ABSTRACT This article discusses a safe and collision-free decision for an autonomous car to make when passing another vehicle on a congested two-way roadway. The proposed algorithm considers both the car going the other way and the vehicle in front of the "ego" car that could be an obstacle. The LIDAR, the medium-range radar, and the long-range radar are the three types of sensors that the ego car will use to observe its surroundings. The second sensor's job is to keep an eye on the vehicle blocking the ego lane. Once that car is detected, the ego car will start to pass it, but only if the first sensor has found that the obstruction has been removed. With the help of the third sensor, the ego car can determine if there is a vehicle approaching from the other direction. After that, the algorithm will determine the time needed to pass both automobiles and their final positions and the distance between them. If there is enough space, the ego vehicle will switch lanes and perform the lane-change movements. If there is not enough space, the ego car will continue driving in the same lane. MATLAB simulation confirms the method, which demonstrates significant improvement. Key words: Autonomous vehicle, overtaking, LIDAR, Radar, 1. INTRODUCTION A self-driving car is capable of autonomous navigation, entirely driven with no human intervention, as they can navigate themselves by recognizing the environment [1], [2]. The development of autonomous vehicle technology shows a rapid increase over time. Researchers and developers have made a lot of progress in detecting obstacles, making navigational decisions, and planning routes, paths, and trajectories. However, if the vehicles around them were evaluated quantitatively, they might be able to figure out the possible risks to a safe and reliable trajectory for self-driving cars.[3]. One of the more significant aspects and concerns of these self-driving cars are overtaking. It is the most frequently used and challenging maneuver for autonomous vehicles apart from the inherent danger because of the difficulty in assessing the required space to perform safe driving. It is under great focus because of its capabilities toward the goal of full end-to-end autonomy. It includes a combination of lateral and longitudinal motion while avoiding collisions and several sub-maneuvers. Extreme care is necessary to handle it since it encompasses risk on single and multi lanes. [4],[5]. One of these researches dealing with overtaking is [6], which used a Stochastic Model Predictive Control (SMPC) to track the desired motion. An adopted probabilistic prediction model for a safe distance to cope with the limitation of cognitive range to perform overtaking. It overtook the surrounding vehicles following traffic regulations and achieved a safe and comfortable overtaking maneuver because of the consideration of vehicles appearing outside the sensor range. Likewise, the overtaking system of [7] uses the current relative position and orientation concerning the overtaken vehicle to decide when to overtake. The authors used feedback control from onboard sensors and applied standard robotic nomenclature for translational and rotational displacements and velocities. The overtaking maneuvers assure a smooth transition between the adjacent phases without using any roadway marking scheme or intervehicle communication. Furthermore, the authors of [5] generate a local risk map, identify a safe target, and plan a trajectory using a modular control framework via MPC controller for autonomous high-speed overtaking. Their study can generate trajectories compatible with vehicle dynamics and safety considerations. The system makes sure that the trajectory, speeding up or slowing down, and moving sideways are possible and can be done in real-time. The studies discussed above contribute significantly and provide an optimal solution for overtaking a slow-moving vehicle for autonomous trajectory planning or lane-change maneuver. However, the said solutions were investigated on one-direction lanes only. There are some autonomous overtaking researches on a two-direction road. One of these is [8], where overtaking takes place in a two-direction high-speed lane. The authors do overtaking by taking a sample of the relative distance to the vehicle in front, getting the inverse of its speed, and using input-output linearization to make the dynamics linear. The formula lets the self-driving car decide when and how to pass another car, even if that other car is in the same or the next lane.Similarly, the automatic driving system of [9] can Lane-Change Maneuver on Overtaking Moving Vehicle on a Two-Way Street Maria Rosario Garcia 1 , Argel Bandala 2 , Ryan Rhay Vicerra 3 1 De La Salle University/Universidad De Manila, Philippines, maria_rosario_t_garcia@dlsu.edu.ph 2 De La Salle University, Philippines, argel.bandala@dlsu.edu.ph 3 De La Salle University, Philippines, ryan.vicerra@dlsu.edu.ph Received Date : April 16, 2022 Accepted Date : May 18, 2022 Published Date : June 06, 2022 ISSN 2278-3091 Volume 11, No.3, May - June 2022 International Journal of Advanced Trends in Computer Science and Engineering Available Online at http://www.warse.org/IJATCSE/static/pdf/file/ijatcse081132022.pdf https://doi.org/10.30534/ijatcse/2022/081132022