Automated Guided Vehicles Dispatching Research Articles

A Self-Attention-Based Deep Reinforcement Learning Approach for AGV Dispatching Systems.

The automated guided vehicle (AGV) dispatching problem is to develop a rule to assign transportation tasks to certain vehicles. This article proposes a new deep reinforcement learning approach with a self-attention mechanism to dynamically dispatch the tasks to AGV. The AGV dispatching system is modeled as a less complicated Markov decision process (MDP) using vehicle-initiated rules to dispatch a workcenter to an idle AGV. In order to deal with the highly dynamical environment, the self-attention mechanism is introduced to calculate the importance of different information. The invalid action masking technique is performed to alleviate false actions. A multimodal structure is employed to mix the features of various sources. Comparative experiments are performed to show the effectiveness of the proposed method. The properties of the learned policies are also investigated under different environment settings. It is discovered that the policies explore and learn the properties of different systems, and also smooth the traffic congestion. Under certain environment settings, the policy converges to a heuristic rule that assigns the idle AGV to the workcenter with the shortest queue length, which shows the adaptiveness of the proposed method.

Automated guided vehicle dispatching and routing integration via digital twin with deep reinforcement learning

The manufacturing industry has witnessed a significant shift towards high flexibility and adaptability, driven by personalized demands. However, automated guided vehicle (AGV) dispatching optimization is still challenging when considering AGV routing with the spatial-temporal and kinematics constraints in intelligent production logistics systems, limiting the evolving industry applications. Against this backdrop, this paper presents a digital twin (DT)-enhanced deep reinforcement learning-based optimization framework to integrate AGV dispatching and routing at both horizontal and vertical levels. First, the proposed framework leverages a digital twin model of the shop floor to provide a simulation environment that closely mimics the actual manufacturing process, enabling the AGV dispatching agent to be trained in a realistic setting, thus reducing the risk of finding unrealistic solutions under specific shop-floor settings and preventing time-consuming trial-and-error processes. Then, the AGV dispatching with the routing problem is modeled as a Markov Decision Process to optimize tardiness and energy consumption. An improved dueling double deep Q network algorithm with count-based exploration is developed to learn a better-dispatching policy by interacting with the high-fidelity DT model that integrates a static path planning agent using A* and a dynamic collision avoidance agent using a deep deterministic policy gradient to prevent the congestion and deadlock. Experimental results show that our method outperforms four state-of-the-art methods with shorter tardiness, lower energy consumption, and better stability. The proposed method provides significant potential to utilize the digital twin and reinforcement learning in the decision-making and optimization of manufacturing processes.

AGV dispatching and bidirectional conflict-free routing problem in automated container terminal

Automated guided vehicles (AGVs) are automated devices that replace internal trucks in container terminals for horizontal container transportation. Compared to internal trucks, AGV has the special characteristic of bidirectional transportation. This characteristic makes AGV change the transport direction without turning around, significantly saving the transport space and time taken up by vehicle transport. However, few studies focus on the bidirectional transportation mode in the AGV scheduling problem to improve the overall efficiency of the automated container terminal (ACT). This study aims at the AGV dispatching and bidirectional conflict-free routing problem. A bi-level mixed integer programming model is proposed, which fully considers equipment coordination, bidirectional conflict-free routing, and import and export container tasks. The bi-level differential evolution algorithm based on the dynamic routing approach is developed to minimize the completion time of all tasks. A series of numerical experiments are conducted to validate the performance of the proposed bidirectional transportation mode and algorithm. The result shows that the efficiency of the bidirectional transportation mode of AGVs is more suitable in the actual ACT operation scenario, and the proposed algorithm can find a better solution within a reasonable amount of time.

A branch-and-bound approach for AGV dispatching and routing problems in automated container terminals

The efficiency of automated container terminals primarily depends on the automated guided vehicle (AGV) scheduling scheme. This paper investigates the AGV dispatching and routing problem with multiple bidirectional paths to generate conflict-free routes. A mixed-integer programming (MIP) model is formulated to minimize the completion time of all jobs (i.e., minimize the makespan). A tailored branch-and-bound (B&B) algorithm is developed to solve the problem within a reasonable amount of time, where the B&B process is used to obtain job assignments and job sequences, and a heuristic algorithm is proposed to generate conflict-free routes. To expedite the proposed algorithm, some accelerating techniques are designed in view of the characteristics of the problem. Computational experiments indicate that the proposed algorithm can obtain near-optimal solutions in small-scale instances, and in large-scale instances, the B&B algorithm significantly outperforms the port scheduling strategies. The proposed algorithm can be applied to real-world cases and dynamic cases, and a two-stage greedy heuristic algorithm is also developed to obtain a satisfactory solution in a short amount of time. Some managerial insights into AGV scheduling are offered to guide terminal operations in this study.

Hybrid multiobjective genetic algorithms for integrated dynamic scheduling and routing of jobs and automated-guided vehicle (AGV) in flexible manufacturing systems (FMS) environment

The paper presents an algorithm for integrated scheduling, dispatching, and conflict-free routing of jobs and AGVs in FMS environment using a hybrid genetic algorithm. The algorithm generates an integrated schedule and detail routing paths while optimizing makespan, AGV travel time, and penalty cost due to jobs tardiness and delay as a result of conflict avoidance. The multi-objective fitness function use adaptive weight approach to assign weights to each objective for every generation based on objective improvement performance. Fuzzy expert system is used to control genetic operators using the overall population performance improvements of the last two previous generations. Computational experiments was conducted on the developed algorithm coded in Matlab to test the effectiveness of the algorithm. Integrated scheduling of jobs in FMS which are in synchrony with AGV dispatching, scheduling, and routing proved to ensure the feasibility and effectiveness of all the solutions of the integrated constituent elements.

Approaches of fuzzy systems applied to an AGV dispatching system in a FMS

Excellence in manufacturing systems has been recognized as one of the main factors behind the success of industrial companies or production companies. New technology for manufacturing processes plays a significant role in this process. Achieving the potential of technological innovations in production, however, requires a wide range of management, as well as engineering issues related to the system. The handling capacity of advanced materials is essential because without this ability of providing the material needed for the proper workstation at the right time and in the right amount, the whole plant will become “bogged down.” This makes it less efficient and thus produces less profit, and/or it operates at higher costs. This paper proposes two approaches for the dispatching of AGV (automated guided vehicle) using systems fuzzy. The first use hierarchical fuzzy rule-based model building in which the main feature is to make the base of fuzzy rules is smaller and simpler but with high coverage and interpretability. The second use adaptive genetic fuzzy system with simple prediction in which the main feature is to increase the sensitivity of the system about the variables. Both approaches using multiple attributes and having the objective decrease the makespan in a FMS (flexible manufacturing system).

자동화 컨테이너 터미널의 복수 규칙 기반 AGV 배차전략 최적화

본 논문은 컨테이너 운송을 위한 AGV(Automated Guided Vehicle) 배차 전략을 대상으로 한다. AGV 배차 문제는 안벽 크레인의 대기 시간과 AGV의 주행 거리를 최소화하도록 AGV에 작업을 할당하는 것이 목표이다. 터미널 환경의 동적인 특성으로 인해 계획 결과의 정확한 예측이 어렵고 수정이 빈번하기 때문에 실무에서는 의사결정 시간이 짧은 단순 규칙 기반 배차가 많이 쓰인다. 그러나 단순 규칙 기반 배차는 근시안적 특성으로 인해 배차의 다양한 성능 지표를 만족시키지 못하는 한계가 있으며 이를 극복하기 위해 본 논문에서는 복수 규칙 기반의 배차 전략을 제안한다. 복수 휴리스틱 기반 배차 전략은 여러 규칙의 가중합으로 구성되며 규칙 사이의 가중치를 최적화하기 위해 다목적 진화 알고리즘을 적용하였다. 시뮬레이션 실험을 통해 제안 방안이 기존 단일 규칙 기반 배차에 비해 더 좋은 성능을 보임을 확인하였다. This paper focuses on dispatching strategy for AGVs(Automated Guided Vehicle). The goal of AGV dispatching is assigning AGVs to requested job to minimizing the delay of QCs and the travel distance of AGVs. Due to the high dynamic nature of container terminal environment, the effect of dispatching is hard to predict thus it leads to frequent modification of dispatching decisions. In this situation, approaches based on a single rule are widely used due to its simplicity and small computational cost. However, these approaches have a limitation that cannot guarantee a satisfactory performance for the various performance measures. In this paper, dispatching strategy based on multicriteria heuristics is proposed. The Proposed strategy consists of multiple decision criteria. A multi-objective evolutionary algorithm is applied to optimize weights of those criteria. The result of simulation experiment shows that the proposed approach outperforms single rule-based dispatching approaches.

Network model and effective evolutionary approach for AGV dispatching in manufacturing system

Automated guided vehicles (AGVs), are the state-of-the-art, and are often used to facilitate automatic storage and retrieval systems (AS/RS). In this paper, we focus on the dispatching of AGVs in a flexible manufacturing system (FMS). A FMS environment requires a flexible and adaptable material handling system. We model an AGV system by using network structure. This network model of an AGV dispatching has simplexes decision variables with considering most AGV problem’s constraints, for example capacity of AGVs, precedence constraints among the processes, deadlock control. Furthermore, these problems can be solved by using a lot of heuristic algorithms as network optimization problems. We are also proposed an effective evolutionary approach for solving a kind of AGV’s problems in which minimizing time required to complete all jobs (i.e. makespan) and minimizing the number of AGVs, simultaneously. For applying an evolutionary approach for this multicriteria case of AGV problem, priority-based encoding method and Interactive Adaptive-weight GA (i-awGA) were proposed. Numerical analyses for case study show the effectiveness of proposed approach.

Strategies for dispatching AGVs at automated seaport container terminals

Control of logistics operations at container terminals is an extremely complex task, especially if automated guided vehicles (AGVs) are employed. In AGV dispatching, the stochastic nature of the handling systems must be taken into account. For instance, handling times of quay and stacking cranes as well as release times of transportation orders are not exactly known in advance. We present a simulation study of AGV dispatching strategies in a seaport container terminal, where AGVs can be used in single or dual-carrier mode. The latter allows transporting two small-sized (20 ft) or one large-sized (40 ft) container at a time, while in single-mode only one container is loaded onto the AGV irrespective of the size of the container. In our investigation, a typical on-line dispatching strategy adopted from flexible manufacturing systems is compared with a more sophisticated, pattern-based off-line heuristic. The performance of the dispatching strategies is evaluated using a scalable simulation model. The design of the experimental study reflects conditions which are typical of a real automated terminal environment. Major experimental factors are the size of the terminal and the degree of stochastic variations. Results of the simulation study reveal that the pattern-based off-line heuristic proposed by the authors clearly outperforms its on-line counterpart. For the most realistic scenario investigated, a deviation from a lower bound of less than 5% is achieved when the dual-load capability of the AGVs is utilized.

Object oriented modelling and development of a dispatching algorithm for automated guided vehicles

The main problem for an automated guided vehicle (AGV) dispatching system is to assign vehicles to transport demands which optimise some predetermined objectives of a manufacturing shop. This paper presents a framework for an AGV dispatching system based on an object oriented approach using the unified modelling language (UML), and the development of a dispatching algorithm to facilitate a human controller to dispatch efficiently a fleet of AGVs in response to calls from any shop floor (or machine) operator. The main reason for this work is to model an AGV dispatching system as well as to develop a dispatching algorithm which can record details of the AGV position and movement and ensure their allocation of new orders. The provision of both immediate and pre-booked orders for an AGV is also incorporated in the proposed AGVs dispatching system. The underlying AGV dispatching system and algorithm are capable of dispatching a vehicle automatically to handle a call at the required time. In order to overcome difficulties associated with tackling immediate orders, pre-booked orders, and processing of information related to AGVs, a comprehensive dispatching algorithm is developed which aims to minimise lateness, traveling time and distance of empty vehicles in a simulated job-shop scenario. The effectiveness of the proposed framework for an AGV dispatching system is shown through a test problem.

A PETRI NET BASED APPROACH FOR AGV DISPATCH SCHEDULING AND FLEET SIZE DETERMINATION

In machine job shops with automated guided vehicles (AGVs) as mean of transportation of job parts, it's usually hard to determine the optimal fleet size of AGVs for achieving a desired performance. In this paper a heuristic AGV dispatching algorithm is constructed based on the concept of ideal pipeline and Petri Net theory to balance the job part scheduling and AGV dispatching that minimize the number of AGVs for achieving zero idle time of the machine. The problem is formulated as polynomial equations owing to the mathematical property of Petri Net and since obtaining the optimal solution is computationally difficult, a way to circumvent the difficulty is necessary for the real application of the algorithm. A heuristic dispatching algorithm that schedules the dispatch of AGVs on a selected window basis is used to reduce the computation time and makes the dispatching algorithm capable of real time implementation.

A value-added approach for automated guided vehicle task assignment

Abstract Efficient material flow is essential in achieving high productivity levels. Design and control of the material handling system directly affect the efficiency of the material flow. In the case of automated guided vehicle (AGV) systems, this translates to guidepath design, AGV dispatching, scheduling, and routing. The focus of this paper is on vehicle dispatching, or vehicle task assignment, which is defined as the selection of the next load for pickup and delivery. The algorithm developed in this research is inherently a demand-driven strategy (that is, pull), which further prioritizes the loads requiring an AGV according to the value added to them as they go through the manufacturing processes. Benchmarking for several performance factors, such as throughput and flow time, through simulation experiments shows this algorithm to outperform some of the best dispatching rules reported in the literature.

Push and pull rules for dispatching automated guided vehicles in a flexible manufacturing system

Abstract Automated guided vehicle (AGV) systems are widely used in flexible manufacturing systems (FMSs) for material handling purposes. Although the AGV systems have provided high flexibility, the design issues on AGV dispatching rules are still to be resolved. The AGV-dispatching rules in an FMS are generally based either on a push or a pull concept. A Petri-net-based simulation was used to investigate the effect of different dispatching rules on the FMS performance. The developed simulation model consists of two modules: a Petri net model and an AGV dispatcher. The two modules are integrated so that the AGV dispatcher controls AGV tokens in the Petri net model. It was shown that there was no significant difference in output rate between push- and pull-based AGV-dispatching rules for a busy FMS.