In this paper, we explore multiple metaheuristic optimization techniques for multi-cooperative production scheduling in the food industry. Our approach aims to minimize production costs by optimizing energy consumption, labor deployment, and raw material utilization while adhering to practical constraints such as demand satisfaction, production line exclusivity, where each line can produce only one product at a time, and line changeover time. Simulated Annealing (SA), Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Social Spider Optimization (SSO) algorithms are implemented and evaluated on this problem. We validate the performance of each algorithm using two case studies, each representing contrasting scenarios in terms of demand and power consumption. Finally, all the results were analyzed and compared, highlighting the PSO algorithm’s efficiency in generating cost-effective schedules compared to other algorithms. However, it came to be very computationally expensive. This study also demonstrates the potential of SSO in addressing complex production scheduling problems in manufacturing, as the results are similar to PSO with a slight difference.

This paper proposes a novel Reinforcement Learning (RL) optimization technique for Road-Side-Units (RSUs) placement along a highway. With RSUs playing a crucial role in Intelligent Transportation Systems (ITS), optimizing RSU placement is crucial for efficient communication. This research aims to optimize the cost of the RSU deployment along the highway and minimize the delay in communication between vehicles. For the purpose of this research, two different deterministic RL algorithms were tested, the Deep Deterministic Policy Gradient (DDPG) and the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithms. Both algorithms are model-free off-policy Actor-Critic algorithms with deterministic actions. The algorithms’ performance was then compared to three different famous Meta-Heuristic algorithms, Simulated Annealing (SA), Genetic Algorithm (GA), and Discrete Particle Swarm Optimization (DPSO). Training results and Key Performance Indicators (KPIs) show a faster run time for the deterministic RL algorithms for the same number of iterations as the Meta-Heuristic algorithms. KPIs also showcase the better solution cost achieved by the RL agents for the specific optimization problem.

In this work, the problem of predicting a pedestrian’s intention to cross the road is addressed using visual data from a camera. The proposed ROS-based modular architecture consists of four modules: Visual-Perception, Intention Prediction, and Planning and Control Modules. The Visual Perception module is divided into three sub-modules. The pedestrian detection is responsible for detecting the pedestrian and analyzing his motion and looking states. The lane detection is responsible for analyzing the structured environment which helps in the road state classifiers. The third sub-module aims to extract some curvilinear localization states that are essential for the vehicle’s motion planning and control. The intention prediction module captures the pedestrian’s intention to cross the road. A comparative study is conducted between three different data-driven sequential models. Each model is trained on the JAAD dataset and different extracted features from the visual perception module. The proposed GRU model obtained an 86% average f1-score, anticipating the pedestrian’s intention two seconds in advance when the pedestrian is standing, and three seconds in advance when the pedestrian is walking to the crossing area. To control the maneuver of the vehicle, longitudinal velocity and lateral controllers are implemented to control the motion of the vehicle while avoid collision with the pedestrian based on the intention prediction. Finally, this work is verified on a 1:4 scaled real vehicle to ensure the applicability of implementing this work in real hardware.

In this paper, we present an innovative indoor positioning system designed for computing the position and orientation of multiple model-scale vehicles. Our system, equipped with four cameras, utilizes a novel image-stitching technique to stitch the images from these cameras into a single output image. We explore diverse methodologies for detecting and estimating the position and orientation of the model-scale vehicles, demonstrating robust detection within the plane. Through rigorous testing with vehicles of varying dimensions, our system achieves an accuracy ranging from approximately 0.5cm to 3cm on both axes. This establishes the system’s potential applicability in scenarios involving model-scale autonomous vehicles, where precise knowledge of each vehicle’s position is critical.

Abstract This paper proposes a solution to enhance and compare different neural network (NN)-based side-slip angle estimators. The feed-forward neural networks (FFNNs), recurrent neural networks, long short-term memory units (LSTMs), and gated recurrent units are investigated. However, there is a lack in the selection criteria of the architectures’ hyper-parameters. Therefore, the genetic algorithm is integrated with the NN-based estimators to find the optimal hyper-parameters for the studied architectures. The tuned hyper-parameters in this work include the number of neurons, number of layers, activation function, optimizer type, and learning rate. The objective function of the optimization problem is minimizing the root-mean-square error (RMSE) on multiple testing data. The optimal models are further included in the design of a hybrid NN estimator with Kalman filter. In the hybrid estimators, the optimal NN estimators are used as virtual sensors to correct the prediction of the side-slip angle resulting from the mathematical lateral vehicle model. Eventually, the performance of the best selected model is evaluated in terms of different metrics; mean RMSE, mean error variance, mean training time, and mean estimation time. LSTMs are found to achieve the lowest mean RMSE while being tested on highly generalized data yielding the highest training and estimation time. However, FFNNs achieve the lowest RMSE while being tested on low generalized data and the lowest training and estimation time. Meanwhile, it is observed that the hybrid estimators achieved lower RMSE with great enhancement compared to the non-hybridized ones proving the effectiveness of the proposed approach and increasing the side-slip estimation generalization ability in unknown environments with high uncertainties, which are not covered by the training dataset for the NNs estimators.

In light of the latest pandemic, many researchers proposed a nature-inspired Coronavirus algorithm to aid in tackling different optimization problems. Many meta-heuristic techniques were conducted to handle the trajectory planning of vehicles on track; however, no implementation or comparison is formulated using the aforementioned new technique. This paper proposes a modified version of the Coronavirus optimization algorithm to aid the drivers offline prior to entering the track. To ensure this algorithm's capability, a comparative study is conducted between the extended Coronavirus Optimization Algorithm (CVOA) and other meta-heuristic techniques such as: Genetic Algorithm (GA) and Grey Wolf Optimization Algorithm (GWO). Using MATLAB, two different tracks were visualized and the aforementioned techniques were employed to generate a feasible trajectory plan. All equations are represented using a real vehicle model to verify the algorithms' outputs with real case scenarios.

Vehicle platooning has become a topic of substantial interest for development of safer and more efficient means of transportation. It is described as a string of connected autonomous vehicles traveling closely, maintaining certain inter-vehicular distances at a set speed boosting the road capacity, improving safety and lowering adverse environmental significance. In this study, several platooning concepts addressed in previous literature are revealed. A unified integrated platoon control architecture is constructed, employing the aforementioned concepts. This architecture contemplates a generic decoupled longitudinal and lateral engine-based vehicle model incorporating the powertrain dynamics. Several control algorithms are investigated, optimized and compared to manage the longitudinal inter-vehicular spacing distances of a formed platoon, as well as the lateral platoon motion tracking. The superior longitudinal controller in terms of spacing error convergence, velocity tracking and acceptable control effort, in addition to the outperformed lateral controller for its accurate lane change tracking and lower computational cost are promoted to be tested in the integrated architecture. Simulations are visualized using the Multi-Robot Systems Intelligent Transportation Systems (MRS-ITS) visualization tool for further realization of the results.

The Connected and Automated Vehicles (CAVs) technology is in continuous growth, especially during the last decade. Accordingly, the development of the V2X protocols grasps the attention of many researchers. However, there is a gap in structuring a holistic architecture that considers the individual agent modules in the transportation system along with their cooperation, especially in the decision-making layer. Thus, the main contribution of this article is to build a Robotics Operating System (ROS)-based architecture, simulating the interaction between the infrastructure and the car agents. The architecture is designed with some essential characteristics: configurable, modular, comprehensive, usable, and generic. In the architecture, the car agent is built to include four main modules to accomplish a high level of autonomy; Localization, Planning, Control, and communication. Meanwhile, the road agent is composed of two modules; mapping and communication. Both agents are considered ROS nodes that can communicate through custom ROS messages through Service/Client architecture using developed V2I protocols. These protocols are responsible for registering the car while joining the road and assigning a unique identifier to each joined car. Moreover, the road assigns the desired speed suiting the road profile and a lane to be kept by the car. The architecture is verified on a designed track map on Webots simulator. A case study of 10 heterogeneous cars is demonstrated to observe the architecture performance. The architecture showed promising results, successfully controlling the cars along with the designed map with acceptable error via the V2I protocols, allowing a human-like driving experience.

This paper addresses the problem of sensors lack that can directly measure the side-slip angle of a vehicle which is crucial in controlling autonomous vehicles. Therefore, estimating the side-slip angle is considered to be the most sufficient and cheapest solution. The main aim is to compare the performance of different Neural Network NN-based estimators through comprehensive performance assessment criteria. This can be done if a proper architecture is designed to allow the adaptation of the estimator to suit any tested data. Four NN approaches are investigated in this work; Feed-Forward Neural Networks (FFNN), Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM) units, and Gated Recurrent Units (GRU). The performance of the approaches is compared to measure the system accuracy represented by Root Mean Square Error and the error variance. Additionally, the computational effort of the architecture is evaluated in terms of the training time and the estimation time. Moreover, the ability of the architecture adaptation to any uncertainty in the validation data is addressed by tuning the network hyper-parameters. IPG CarMaker aided the evaluation process, by providing different roads and cars, that nearly replicates real-life conditions. The tested data show promising results in terms of accuracy, computation effort, and the robustness of the architecture. Regarding the performance assessment, it is shown that the FFNN achieves higher accuracy compared to the NN with recurrence property. Meanwhile, the GRUs surpassed the FFNN in terms of mean training time.

Over the years, it is observed that the growth rate of the Autonomous Vehicles' (AVs) field is boosting. Nevertheless, AVs research still faces a gap in understanding human behaviour. That's why recent studies focused on predicting the intentions of pedestrians in urban traffic scenes. Most of the recent works dealt with the intention prediction problem as a trajectory prediction problem. However, focusing only on the trajectory ignores the context of the scene and the pedestrian's behaviour. Also, if the pedestrian is in a standing state, it will be hard to predict his intention using trajectory prediction. In this study, the problem of crossing intention prediction of pedestrians is addressed using Spatio-temporal visual data. Moreover, this work focuses on predicting the intentions of one pedestrian and obtaining the best results in the least computational time as this work is a pre-stage for hardware implementation in the future. The proposed architecture consists of three phases starting with the detection phase (phase one), extraction phase (phase two) and ending by the prediction phase (phase three). The first two phases obtain the spatial dependency by detecting the pedestrian and analysing his state (Walking/Standing, LookingINot-Looking) from a monocular RGB camera. The third phase obtains the temporal dependency by adding the sense of time through connecting a sequence of consecutive data obtained from the two previous phases. During the prediction phase, a comparative study is conducted between three different data-driven sequential models. Each model is trained on the JAAD dataset and the considered inputs are the motion state, looking state, and whether the pedestrian is on-road or on the curb. The criteria for choosing the best model is based on compromising between the Fl-score and model computational time. For the third phase, the chosen model, based on the indicated criteria, is the Gated Recurrent Unit (GRU) model. Finally, the proposed architecture reached a performance of 12 FPS (Frames per second) with 79% average F1-score and 88 % average precision score, and can predict pedestrian's intention three seconds before crossing.

As a result of Demand Response (DR) programs implementation in the industrial sector varying electricity prices based on Time-of-Use (ToU) rates are becoming more common replacing traditional flate-rates per unit of energy consumption. On the other hand increased automation of industrial facilities is gaining interest due to their reliability flexibility and robustness. However it is necessary to determine a suitable task schedule in order to ensure their cost-efficiency and maximize profits. In this study a Market-Based approach is considered to solve the Multi-Agent Task Allocation (MATA) problem for a group of homogeneous agents and tasks. While most previous studies model the problem considering flate-rates for electricity consumption the main contribution of this study is accounting for the implementation of a DR program with varying ToU rates. The effects of optimizing the task allocation process on the costs incurred are investigated and compared to the effects of random assignment. Four different case studies are analyzed considering different-sized maps and number of tasks. The results show the computational efficiency of the proposed algorithm and its ability to massively decrease the electrical charging costs.

Cooperative driving has caught the interest of many research centers around the world as it introduced a solution to many problems in traffic and especially the reduction of accidents. One of the cooperative driving applications is vehicle platooning, which has proven its ability to increase road capacity and decrease fuel consumption. As the platoon has to have a rules of organization to form and split safely, this paper introduces a dynamic platoon model and protocols with different control techniques applied on different sub-systems in the model. Additionally, the basic platooning maneuvers are governed by applying certain protocols to organize the formation of platoon in a safe manner. Several tests on the platooning model and maneuvering controls have been performed from real-life scenarios and using real-vehicle data and a software in the loop to verify the results and evaluate the performance of the proposed approaches. The results prove an acceptable performance of the control techniques applied on the platooning model.

State estimation is one of the essential tasks for autonomous systems, which is required in multiple applications, such as localization, object tracking, mapping, and many more. Various solution paradigms have been proposed by the scientific community to solve this well established problem. Some of the commonly used techniques are based on Bayesian inference, specifically the particle filter as it has the advantage of modeling arbitrary distributions without the unimodal assumption limitation. The particle filter has proven to be successful in many applications, however, a large set of particles is required in order to have robust estimates against noisy measurements and erroneous data association, which impedes the runtime performance of the filter. This study shows that a learning framework that trains a recurrent neural network with labeled data generated from the probabilistic estimation of a particle filter is capable of exploiting the expressive power of neural networks in order to capture the behavior of the filter and handle the noisy sensory information. The trained model is capable of performing the estimation with lower runtime complexity, making it applicable to numerous autonomous systems. This approach also proves to be helpful in situations when the ground truth data can not be accessed. We train and evaluate our strategy using raw GPS sensor measurements from the Oxford RobotCar dataset. The results show that the performance of the recurrent network closely matches that of the particle filter without exhaustive tuning and that the network is able to generalize effectively on test datasets as well.

Practical experimentation and observations play an essential role in the experience of learning. And several emerging remote-labs are making the best use of the internet to achieve a better experience for the students through providing several real experiments from a website. The field of biotechnology involves several experiments that involves living cells manipulation for analysis and understanding. In this study, an adaptive living cell classifier is proposed through the remote e-lab technology. The classifier is able to distinguish between several types of cells. Furthermore, the classifier continuously learns through utilizing each new input image to train and improve its adaptive-neuro fuzzy (ANFIS) classifier for better decisions in the future. The results achieved from the proposed system are promising and opens the door for further experiments implementation along this field.