Left Turn Research Articles

A Deep Reinforcement Learning Agent for Snake Game

After watching AlphaGo a Netflix documentary which presents how AlphaGo is an AI computer game developed by deep-mind technologies based on deep reinforcement learning (DRL). Since then, my interest in reinforcement learning has been growing. In this project, I will apply reinforcement learning to develop an agent to play snake game. Where Deep learning will implement a neural Network to help the agent (snake) to learn what action must take to get a state. If we describe deep reinforcement learning (DRL) model where agent interacts with an environment and chooses an action. Based on action, agents receive feedback from the environment as states (or perceives) and rewards. A state = an array with 11 input values, each input values represent a neural network that provides an output of 3 values, each one represents three possible actions the agent (snake) can take (Straight, Right Turn and Left Turn).

Optimization Study of Non-motorized left Turn at Intersections based on Webster's Timing Method

Lhasa tourism drives Lhasa's economic development and also brings a considerable burden to Lhasa's transportation, so solving the problem of traffic congestion caused by the increase of non-motorized traffic flow at the intersection in Lhasa is of great significance to the development of Lhasa's transportation. In order to solve the problem of non-motorized vehicles turning left at the intersection and alleviate the traffic pressure in Lhasa, this paper constructs a multi-phase signal timing optimization scheme based on the Webster timing method model, and takes the intersection of Dangre Middle Road and Niangre Middle Road in Lhasa as an example, and through the simulation of Vissim simulation and delay calculation analysis, it is concluded that the multi-phase signal timing model constructed in this paper is better than the traditional four-phase signal timing in terms of the traffic organization behavior of the left-turning of non-motorized vehicles. Traditional four-phase signal timing, this paper on the six-phase signal timing scheme research on the future intersection of non-motorized traffic capacity and the intersection of the level of service to improve the multi-phase signal timing scheme to provide a multi-phase signal timing scheme optimization ideas.

Application of machine learning models to predict driver left turn destination lane choice behavior at urban intersections

When there are multiple lanes to choose from downstream of a turning movement, drivers should choose the innermost lane so that drivers at other approaches of the intersection may make concurrent turning movements in the outermost lane(s). However, human drivers do not always choose the innermost lane, which could lead to crashes with other vehicles. Therefore, predicting human driver behaviors is vital in reducing crashes, as the need to share the roadways with automated vehicles (AVs) continues to grow. In this research, various machine learning models have been used to predict the left turn destination lane choice of human-driven vehicles (HDVs) at urban intersections based on several quantifiable parameters. A total of 174 subject vehicles were extracted and analyzed in Los Angeles, California, and Atlanta, Georgia, using HDV trajectory data from the Next Generation SIMulation (NGSIM) database. Five machine learning techniques, namely binary logistic regression, k nearest neighbors, support vector machines, random forest, and adaptive neuro-fuzzy inference system, were applied to the extracted data to predict the lane choice behavior of drivers. The k nearest neighbors model showed the most promising results for the evaluated data with a correct decision score of over 93% for the unseen test data. This model may be programmed into: (i) AVs, in conjunction with sensors, to predict if an HDV is about to turn into the incorrect destination lane; and (ii) microscopic traffic simulation tools so that modelers can identify potential conflicts when HDVs do not select the appropriate destination lane.

Query-Informed Multi-Agent Motion Prediction.

In a dynamic environment, autonomous driving vehicles require accurate decision-making and trajectory planning. To achieve this, autonomous vehicles need to understand their surrounding environment and predict the behavior and future trajectories of other traffic participants. In recent years, vectorization methods have dominated the field of motion prediction due to their ability to capture complex interactions in traffic scenes. However, existing research using vectorization methods for scene encoding often overlooks important physical information about vehicles, such as speed and heading angle, relying solely on displacement to represent the physical attributes of agents. This approach is insufficient for accurate trajectory prediction models. Additionally, agents' future trajectories can be diverse, such as proceeding straight or making left or right turns at intersections. Therefore, the output of trajectory prediction models should be multimodal to account for these variations. Existing research has used multiple regression heads to output future trajectories and confidence, but the results have been suboptimal. To address these issues, we propose QINET, a method for accurate multimodal trajectory prediction for all agents in a scene. In the scene encoding part, we enhance the feature attributes of agent vehicles to better represent the physical information of agents in the scene. Our scene representation also possesses rotational and spatial invariance. In the decoder part, we use cross-attention and induce the generation of multimodal future trajectories by employing a self-learned query matrix. Experimental results demonstrate that QINET achieves state-of-the-art performance on the Argoverse motion prediction benchmark and is capable of fast multimodal trajectory prediction for multiple agents.

Латинская Америка: колебания политического маятника Ивановский З.В., ред. Латинская Америка: политический ландшафт на фоне турбулентности. М.: ИЛА РАН, 2022

The left and right, democratic, hybrid, and authoritarian systems, consolidated and unconsolidated democracies, rising giant-states and underdeveloped economies coexist in the region of Latin America and the Caribbean, which covers more than three dozen states that share common historical destiny, language, and religion. This makes the region a fertile ground for studying various types of regimes, political processes and alternative development paths in the modern world, as vividly demonstrated by the collective monograph “Latin America: Political Landscape in the Midst of Turbulence”, published at the end of 2022. Based on a thorough analysis of the situation in the region, the authors of the monograph show that political life in Latin America is characterized by instability and constant reshaping of the political landscape. The strengthening of the positions of right-wing forces replaced the left turn of the 2000s, and since the beginning of the 2020s, the continent has again drifted to the left. The authors pay special attention to the following factors of turbulence in the development of the region: a natural-resource-based economy in most Latin American countries, poor quality of governance, a decline in trust in political institutions caused by regular corruption scandals, the criminalization of society, the increase in drug trafficking and violence, and the inability of law enforcement agencies to ensure citizens’ safety. Several chapters of the monograph are devoted to the varieties of regimes of the “21st century socialism” that exist in Venezuela, Nicaragua and Bolivia, as well as to the oldest socialist regime on the continent in Cuba. According to the reviewer’s conclusion, with the publication of this monograph the readers received valuable work that allows them to better comprehend the essence of the complex political processes that take place in Latin America today.

Are Digital Markers of Everyday Driving Comparable to Fluid Biomarkers in Identifying Preclinical Alzheimer Disease?

AbstractBackgroundPreclinical Alzheimer’s disease (AD) may have a subtle functional signature reflected in changes in everyday driving behaviour. Driving has the potential to serve as a digital marker for AD when captured continuously and characterized accurately. This study directly compares the diagnostic performance of digital markers of everyday driving, cerebrospinal fluid (CSF), and blood biomarkers in detecting amyloid deposition in AD, using amyloid positron emission tomography (PET) as the “gold standard”.MethodParticipants were enrolled in a longitudinal study on driving and preclinical AD biomarkers at Washington University School of Medicine. The Driving Real‐World In‐Vehicle Evaluation System (DRIVES) Project uses in‐vehicle GPS dataloggers to collect daily driving among cognitive normal drivers. This study included 121 drivers (aged 65+) with complete blood, CSF, and PET biomarker data. We employed three artificial neural network (ANN) to examine the relationship between everyday driving and PET‐amyloid status. The first model solely included driving features. The second ANN model (i.e., “non‐intrusive” model) added age and education level as additional features. The third model included all of the previous features plus APOE e4 status. Plasma Aß42/Aß40<0.1013 and CSF Aß42/Aß40<0.0673 were used to detect PET‐amyloid positivity. To compare the performance of the ANN models and the CSF and plasma biomarkers, area under the receiver operating curve (AUC) from 5‐fold cross‐validation were calculated.ResultIndividuals who were PET‐amyloid positive (n = 46) were more likely to be older (p<.05) and carry an APOE e4 allele (p<.0001). The two groups did not differ in sex or education. In predicting PET‐amyloid positivity, the five most important driving features were the number of left turns per mile, 80th percentile of jerk, minimum vehicle speed, speed variability, and the number of speeding incidents. The CSF biomarker achieved the highest performance (AUC = 0.96±0.04), followed by the model with driving, age, education and APOE e4 status (AUC = 0.88±0.05). The model with only driving features achieved lower performance (AUC = 0.74±0.07), while the non‐intrusive ANN model achieved comparable performance to the plasma biomarker (AUC: 0.82±0.11vs.0.83±0.12).ConclusionDigital markers of everyday driving along with age and education level offer a non‐intrusive and scalable diagnostic tool for preclinical AD that provides ongoing assessments.

IMPROVING MOBILITY IN KEDIRI CITY WITH MKJI ANALYSIS 1997 AND TIME SLICE AT SIMPANG EMPAT DHOHO PLAZA

The city of Kediri, as the third largest city in East Java, has cultural heritage, history and natural beauty that attracts visitors from various regions. However, population growth and motorized vehicles cause traffic jams. Signalized intersections, which are a solution to control intersection conflicts, require knowledge of intersection saturation flows to plan efficient traffic management. Analysis methods such as MKJI 1997 and Time Slice with Vehicle Operating Cost (BOK) calculations help design effective solutions. This research shows that adjusting cycle times, increasing effective width, and optimizing signals can improve transportation system performance and reduce congestion in Kediri City. Cycle time analysis taking into account factors such as vehicle type, traffic volume, and vehicle speed resulted in a cycle time of 90.7 seconds. The impact can be seen in the Degree of Saturation (DS) at various intersection arms, with service variations from E to F. Research also shows a decrease in waiting time/queuing for signal lights from 135 seconds to 118 seconds, which results in a decrease in the degree of saturation (DS) and changes in time cycle. The use of Smart Traffic Lights and traffic engineering, such as immediate left turns on the east arm and additional signs, is required. Evaluation of congestion cost efficiency shows that changing the value of side barriers can save Rp. 22,679,183 in the next five years.

Review on the Development of Left Turn Crossing Theory of Non-Motor Vehicles

Non-motor vehicles are an important part of urban road traffic system, which plays an important role in the daily travel of urban residents. In recent years, the congestion of urban road traffic system has become more and more serious. In order to alleviate the congestion situation, on the one hand, the traffic management department reduces the road congestion by limiting the number of motor vehicles; on the other hand, the government vigorously advocates public transportation and green and low-carbon travel to adjust the travel mode structure. This paper systematically combs, summarizes and summarizes the views of domestic and foreign scholars from two aspects of the mode and theoretical development of non-motor vehicles turning left across the street. It is found that the left-turn crossing behavior of non-motor vehicles is from simple to comprehensive, from immature to gradually mature. Finally, the research on the development of non-motor vehicle holiday behavior is summarized and prospected.

Democracy on the ground: local politics in Latin America's left turn

Democracy on the ground: local politics in Latin America's left turn

Seeking empathy or suggesting a solution? Effects of chatbot messages on service failure recovery

Chatbots as prominent form of conversational agents are increasingly implemented as a user interface for digital customer-firm interactions on digital platforms and electronic markets, but they often fail to deliver suitable responses to user requests. In turn, individuals are left dissatisfied and turn away from chatbots, which harms successful chatbot implementation and ultimately firm’s service performance. Based on the stereotype content model, this paper explores the impact of two universally usable failure recovery messages as a strategy to preserve users’ post-recovery satisfaction and chatbot re-use intentions. Results of three experiments show that chatbot recovery messages have a positive effect on recovery responses, mediated by different elicited social cognitions. In particular, a solution-oriented message elicits stronger competence evaluations, whereas an empathy-seeking message leads to stronger warmth evaluations. The preference for one of these message types over the other depends on failure attribution and failure frequency. This study provides meaningful insights for chatbot technology developers and marketers seeking to understand and improve customer experience with digital conversational agents in a cost-effective way.

Frequency of injuries and orthopaedic conditions sustained by flyball dogs

Abstract The risk of injury to canine flyball competitors has been noted as similar to that reported for canine agility, affecting approximately one third of competitors throughout their careers. To date, no studies have examined whether elements such as consistently turning in one direction during the box turn are associated with specific types of injuries. The aims of this study were to describe the frequency and types of injuries experienced by flyball dogs and to evaluate variables that may affect injury occurrence at specific anatomical sites which may give insight into potential ways of reducing injury risk. An online questionnaire, sent to owners, was used to obtain data on 75 dogs that had sustained an injury whilst training or competing in flyball in the last five years. Among the 75 dogs, the most common location was the shoulder region (n = 17; 16.8% of injured dogs), followed by the back and trunk (n = 16; 15.8%). Back injuries had the highest recorded number of repeat injuries n = 66 (24.63%), and when all individual injuries were considered, 44.40% (n = 119) of them occurred on the forelimbs. The most common type of injury was inflammation to soft tissue. There was a statistically significant association between the direction the dog turns at the box and the side of occurrence of injuries at the shoulder (X2(2) = 13.71, ), with injuries most likely to occur to the limb on the inside of the turn (left turn = left shoulder). For hip injuries, there was also an association with the most affected side and the side of the box turn (X2(2) = 6.702, ), with injuries most likely to occur on the limb at the outside of the turn (left turn = right hip). These findings identify new risk factors for injury within flyball which can be used to improve welfare and safety for competitors.

Modeling and analysis of vehicle path dispersion at signalized intersections using explainable backpropagation neural networks

The dispersion of vehicular paths is a common phenomenon in the inner area of signalized intersections due to heterogeneous driver behavior and interactions. This study aims to develop an explainable neural network-based model to describe the vehicle path dispersion by exploring the relationship between the path dispersion and external factors. A backpropagation neural network model was established to analyze the effects of external factors on the dispersion of through and left-turn paths based on real trajectory data collected from 20 intersections in Shanghai, China. Twelve influencing factors in varying geometric, traffic, signalization, and traffic management conditions were considered. The predictive power and transferability of the model were verified by applying the trained model on the four new intersections. The contributions of the influencing factors on the path dispersion were explored based on the neural interpretation diagram, relative importance of influencing factors, and sensitivity analysis to offer explanatory insights for the proposed model. The results show that the mean absolute percentage errors of the path dispersion models for the through and left-turn movements are only 14.67% and 17.65%, respectively. The through path dispersion is primarily influenced by the number of exit lanes, the offset degree between the approach and exit lanes, and the traffic saturation degree on the through lane. In contrast, the path dispersion of the left turn is mainly affected by the number of exit lanes, the left-turn angle, and the setting of guide lines.

ПОВЫШЕНИЕ ЭФФЕКТИВНОСТИ ОРГАНИЗАЦИИ ДОРОЖНОГО ДВИЖЕНИЯ МЕТОДОМ ОТНЕСЕННОГО ЛЕВОГО ПОВОРОТА

The object of the study is the process of traffic flow in cities, and the subject is this process in relation to regulated street intersections. The main goal of the study is improving the efficiency of traffic organization by preventing the formation of traffic congestion on the street and road network. To achieve this, the creation of an offset left turn was proposed. During the study, the street and road network of Tyumen was considered, and areas for possible construction of the offset left turn were identified. The regulated intersection of 50 Let VLKSM St. and Geologorazvedchikov St. was chosen as the most appropriate. The results of the study showed that the use of the offset left turn would significantly reduce traffic delays and accident risk by reducing conflict points. The evaluation of the effectiveness of proposed solutions also confirmed the positive technological, social, economic and environmental effects.

A thalamo-parietal cortex circuit is critical for place-action coordination.

The anterior and lateral thalamus (ALT) contains head direction cells that signal the directional orientation of an individual within the environment. ALT has direct and indirect connections with the parietal cortex (PC), an area hypothesized to play a role in coordinating viewer-dependent and viewer-independent spatial reference frames. This coordination between reference frames would allow an individual to translate movements toward a desired location from memory. Thus, ALT-PC functional connectivity would be critical for moving toward remembered allocentric locations. This hypothesis was tested in rats with a place-action task that requires associating an appropriate action (left or right turn) with a spatial location. There are four arms, each offset by 90°, positioned around a central starting point. A trial begins in the central starting point. After exiting a pseudorandomly selected arm, the rat had to displace the correct object covering one of two (left versus right) feeding stations to receive a reward. For a pair of arms facing opposite directions, the reward was located on the left, and for the other pair, the reward was located on the right. Thus, each reward location had a different combination of allocentric location and egocentric action. Removal of an object was scored as correct or incorrect. Trials in which the rat did not displace any objects were scored as "no selection" trials. After an object was removed, the rat returned to the center starting position and the maze was reset for the next trial. To investigate the role of the ALT-PC network, muscimol inactivation infusions targeted bilateral PC, bilateral ALT, or the ALT-PC network. Muscimol sessions were counterbalanced and compared to saline sessions within the same animal. All inactivations resulted in decreased accuracy, but only bilateral PC inactivations resulted in increased non selecting, increased errors, and longer latency responses on the remaining trials. Thus, the ALT-PC circuit is critical for linking an action with a spatial location for successful navigation.

A - 86 Using Virtual Reality Driving Simulation to Determine Cognitive Correlates of Turning Behaviors among Drivers with and without Multiple Sclerosis.

This study uses a virtual reality driving simulator (VRDS) to 1) compare turning behavior between adults with or without MS; 2) identify cognitive correlates of turning behavior. 64 community-dwelling adult drivers, including 38 people with MS (PwMS) and 26 healthy controls (HCs) (ages 21-60, 78% women, 69% White), completed neuropsychological and VRDS assessments. Key VRDS turning variables were lane position, speed, acceleration, and braking. Cross-sectional analyses included 1) analysis of variance and chi-square tests; 2) Pearson and Spearman's correlations. HCs took sharper left turns than PwMS (p < 0.05, η2 = 0.08). There were no other between-group differences. Weaker visuospatial functioning [r = -0.301, 95% CI = (-0.512,-0.050), p = 0.018], and slower gait [r = 0.0328, CI = (0.063,0.55), p = 0.014], manual motor speed [r = 0.378, CI = (0.130,0.582), p = 0.03], and psychomotor speed [r = -0.334, CI = (-0.545,-0.077), p = 0.011] were associated with greater lane position variability. Greater relative acceleration before the turn was linked to worse performance in psychomotor speed [r = 0.397, CI = (0.148,0.593), p = 0.002], time-based working memory [r = 0.452, CI = (0.221,0.631), p < 0.001], and visuoperceptual judgment time [r = -0.482, CI = (-0.660,-0.244), p < 0.001]. Slower cognitive/motor speed and poorer visuospatial abilities/memory were associated with slower driving and more braking (all p < 0.05). Drivers with and without MS exhibited similar turning behavior, but PwMS were more cautious during left turns, possibly reflecting a compensatory strategy. Despite few group differences, several cognitive domains commonly affected by MS were associated with turning behaviors. Slower cognitive/motor speed and worse visuospatial abilities were associated with less smooth control of spatial position and less safe turn preparation, but slower driving. Future studies will include larger samples and other clinical populations and may inform clinical driving recommendations.

Improving Pedestrian Progression at Multistage Crossings Using Pedestrian–Left Turn Overlaps

At signalized intersections where pedestrian crossings are divided into two stages by using a median island, pedestrian delay can be very long if the partial-crossing phases have poor progression. One technique for reducing pedestrian delay is pedestrian–left turn overlaps, in which half-crossings that are not in conflict with a left turn are allowed to run during the left turn phase as well as during the parallel vehicle through phase. Such an overlap can enable some pedestrians to cross in a single pass, or to have a far shorter wait at the median. Because this technique requires surrendering some of the flexibility with which left turn phases normally operate—in phases with a pedestrian call, the left turn phase cannot be skipped and may require a longer minimum green—there can also be an impact on vehicle delay. A case study of two intersections in Brookline, MA found that, compared with the current timing plan, using pedestrian–left turn overlaps could reduce pedestrian delay from about 100 to 35 s, with a negligible impact on vehicle delay. A methodological innovation introduced in this study was that average pedestrian delay was calculated for three pedestrian speeds and averaged together, thus accounting for how pedestrian delay at multistage crossings can be far lower for faster pedestrians who may be able to cross in a single pass than for slower pedestrians who have to wait at the crossing island.

Efficient Deep Reinforcement Learning With Imitative Expert Priors for Autonomous Driving.

Deep reinforcement learning (DRL) is a promising way to achieve human-like autonomous driving. However, the low sample efficiency and difficulty of designing reward functions for DRL would hinder its applications in practice. In light of this, this article proposes a novel framework to incorporate human prior knowledge in DRL, in order to improve the sample efficiency and save the effort of designing sophisticated reward functions. Our framework consists of three ingredients, namely, expert demonstration, policy derivation, and RL. In the expert demonstration step, a human expert demonstrates their execution of the task, and their behaviors are stored as state-action pairs. In the policy derivation step, the imitative expert policy is derived using behavioral cloning and uncertainty estimation relying on the demonstration data. In the RL step, the imitative expert policy is utilized to guide the learning of the DRL agent by regularizing the KL divergence between the DRL agent's policy and the imitative expert policy. To validate the proposed method in autonomous driving applications, two simulated urban driving scenarios (unprotected left turn and roundabout) are designed. The strengths of our proposed method are manifested by the training results as our method can not only achieve the best performance but also significantly improve the sample efficiency in comparison with the baseline algorithms (particularly 60% improvement compared with soft actor-critic). In testing conditions, the agent trained by our method obtains the highest success rate and shows diverse and human-like driving behaviors as demonstrated by the human expert. We also find that using the imitative expert policy trained with the ensemble method that estimates both policy and model uncertainties, as well as increasing the training sample size, can result in better training and testing performance, especially for more difficult tasks. As a result, the proposed method has shown its potential to facilitate the applications of DRL-enabled human-like autonomous driving systems in practice. The code and supplementary videos are also provided. [https://mczhi.github.io/Expert-Prior-RL/].

Analysis and Comparative Study of Signalized and Unsignalized Intersection Operations and Energy-Emission Characteristics Based on Real Vehicle Data

With the development of the economy, urban road transportation has been continuously improved, and the number of motor vehicles has also increased significantly, leading to serious energy consumption issues. As critical nodes in the urban road transportation network, intersections have become a focal point of research on vehicle energy consumption. To investigate whether traffic signal lights affect fuel consumption and emissions, this study analyzed the operating characteristics, fuel consumption, and emissions of intersections with and without traffic signal lights using real-world vehicle data. The data from the INTERACTION dataset for both signalized intersection VA and unsignalized intersection MA are used in the study, with a time duration of 3200 s. The VT-micro energy consumption and emissions model was applied to calculate and comprehensively analyze the vehicle flow, fuel consumption, and emissions. Additionally, the study compared the fuel consumption and emissions for different driving scenarios, including straight through, left turn, right turn, and U-turn, within a single traffic signal cycle. The results revealed that at signalized intersections, the average fuel consumption per vehicle was 26.54 L/100 km, NOx emissions were 68.76 g/100 km, and CO2 emissions were 61.07 g/100 km. In contrast, at unsignalized intersections, the average fuel consumption per vehicle was 46.88 L/100 km, NOx emissions were 149.26 g/100 km, and CO2 emissions were 107.16 g/100 km. The study indicated that for traffic volumes between 50 and 103 vehicles per 100 s, signalized intersections demonstrated better fuel consumption and emission performance than unsignalized intersections. Signalized intersections could accommodate larger traffic volumes and provide enhanced traffic safety. In conclusion, the findings of this study are important for urban traffic planning and environmental policies. They provide a scientific basis for reducing fuel consumption and emissions and improving road traffic efficiency. Due to the advantages of signalized intersections in terms of energy consumption and emissions, future urban traffic planning should consider more signal light controls to achieve energy savings, emission reduction, and improved traffic operation efficiency.

Network-Wide Implementation of Roundabouts Versus Signalized Intersections on Urban Streets: Analytical and Simulation Comparison

This paper examines the impact of roundabouts implemented at intersections throughout a dense urban network on its operational performance. Metrics considered include the average free-flow speed, flow-moving capacity, trip-serving capacity, and fuel consumption rate. Three intersection strategies are compared: signalized intersections allowing left turns in a permitted manner (TWs), signalized intersections prohibiting left turns (TWLs), and modern roundabouts (RBs). Using the approaches of macroscopic fundamental diagrams and network exit functions, both analytical investigations and microscopic traffic simulations for grid networks were conducted. In general, the results from both analyses agree well. The results reveal that when single-lane roundabouts are applied in networks with a single travel in each direction, the RB network outperforms the TW network for all operational metrics. The RB network also outperforms the TWL network in free-flow speed and flow-moving capacity and has a similar trip-serving capacity as the TWL network. However, when roundabouts with two travel lanes are applied on multi-lane networks, the TWL network exceeds the RB network in both flow-moving and trip-serving capacities. This decrease in the performance of the RB network could possibly come from the complexity imposed on the entering vehicle that wants to use the inner lane. Moreover, because vehicles in the RB network need to accelerate/decelerate more frequently those in the other networks, the RB network generates a higher fuel consumption rate in uncongested and capacity conditions. The findings suggest intersections of roundabouts could be beneficial for networks with a single travel lane in each direction.

Application of PID Control System in Mecanum Wheelchair

This research centers on the design and implementation of a control system for an electric wheelchair equipped with Mecanum wheels. The study details a comprehensive research methodology starting with the creation of a block diagram to guide system design, hardware selection, and overall implementation. The electric wheelchair system incorporates power resources, input devices, and energy output mechanisms, utilizing a 24 VDC battery and a joystick with a 10K ohm potentiometer connected to an Arduino Due microcontroller. The operational workflow of the system is defined, enabling the wheelchair to respond to joystick commands for forward, left turn, right turn, and other movements. A PID control system is employed to regulate motor movement, enhancing control precision. The Cohen-Coon tuning method is used to determine the PID controller's gain, ensuring efficient closed-loop control. Results from PID controller experiments under P control and PD control are presented, demonstrating the system's responses for different gain values. Optimal performance is observed with a Kp value of 80 and Kd value of 1.2, showcasing improved response speed, reduced rise time, enhanced setting time, and lower percent overshoot. In conclusion, the combined proportional and derivative control system, specifically with Kp = 80 and Kd = 1.2, proves to be effective in enhancing the Mecanum wheelchair's performance. This study provides valuable insights into precise parameter adjustments for optimal control in Mecanum wheelchair applications.