Language models (LMs) possess a strong capability to comprehend natural language, making them effective in translating human instructions into detailed plans for simple robot tasks. Nevertheless, it remains a significant challenge to handle long-horizon tasks, especially in subtask identification and allocation for cooperative heterogeneous robot teams. To address this issue, we propose a Language Model-Driven MultiAgent PDDL Planner (LaMMA-P), a novel multi-agent task planning framework that achieves state-of-the-art performance on long-horizon tasks. LaMMA-P integrates the strengths of the LMs' reasoning capability and the traditional heuristic search planner to achieve a high success rate and efficiency while demonstrating strong generalization across tasks. Additionally, we create MAT-THOR, a comprehensive benchmark that features household tasks with two different levels of complexity based on the AI2-THOR environment. The experimental results demonstrate that LaMMA-P achieves a 105% higher success rate and 36 % higher efficiency than existing LM-based multiagent planners. The experimental videos, code, datasets, and detailed prompts used in each module can be found on the project website: https://lamma-p.github.io.

Many applications require the deployment of legged-robot teams to effectively and efficiently carry out missions. The use of multiple robots allows tasks to be executed concurrently, expediting mission completion. It also enhances resilience by enabling task transfer in case of a robot failure. This paper presents a formulation based on Mixed Integer Linear Programming (MILP) for allocating tasks to robots by taking into account travel time and ensuring efficient execution of collaborative tasks. We extended the MILP formulation to account for complexities with legged robot teams. Our results demonstrate that this approach leads to improved performance in terms of the makespan of the mission. We demonstrate the usefulness of this approach using a case study involving the disinfection of a building consisting of multiple rooms.

Shipborne unmanned aerial vehicles (UAVs) are safer and more flexible for maritime missions, but frequent recharging is needed during long-term patrols. A coordinated system for path planning between ships and electric UAVs is necessary for efficient large-area coverage. A two-stage approach is proposed to minimize the makespan overall UAVs’ flight and the move distance overall ships combinationally. First, the target space is triangularized corresponding to the UAV camera field of view for generating air waypoints. Second, a MILP model is designed to connect suitable air waypoints for UAVs and marine waypoints for ship(s) to form the optimal path for them coordinating the requirements of the area coverage and the UAV recharging. The simulation experiments show the proposed model works for the scenario of either the static or the dynamic motherships in a unified way. In the static mode, the vessels are not migrated and the number of vessels and ship calling points required is the same. In the dynamic model, the ship can be repositioned to recover and recharge the drone, and the task can be accomplished simply by repositioning the ship between waypoints. Dynamic models have better interaction patterns than static models.

The parallel batch processing machine scheduling problem with two-dimensional bin packing constraints (PBS-2D) has appeared in many industrial environments, but the research on this problem is still not sufficient. In this article, we use the 2-D rectangular packing constraints to describe the spatial geometric layout of jobs more accurately, and propose a mixed integer linear programming (MILP) model to minimize the makespan of PBS-2D. According to the in-depth analysis of the problem properties, PBS-2D is separated into three subproblems, which are job-machine assignment, 2-D job placement, and job sequence optimization. An improved biased random key genetic algorithm (OBRKGA) is proposed to solve job-machine assignment and job sequence optimization. This algorithm generates the initial population based on the idea of orthogonal experimental design, so that the whole solution space can be scanned evenly when generating the initial solution, making the algorithm be more likely to find better solutions for subsequent evolution. At the same time, a bin packing algorithm based on best fit is designed to handle 2-D job placement problem. Finally, the algorithms are tested on a large number of randomly generated instances, the results show that OBRKGA outperformed the original BRKGA, the classical moth-flame optimization algorithm, particle swarm optimization algorithm, gravity search algorithm, and MILP.

Spatial resource allocation is common in batch scheduling problems. It is usually modeled as simple capacity constraints, which hurts the accuracy of a solution if a geometric layout is required in multidimensional space. In this article, a multiobjective mixed-integer linear programming model including total weighted waiting time and resource utilization is proposed for single batch processing machine scheduling, in which the two-dimensional (2-D) rectangular packing constraints are introduced to handle a 2-D layout more precisely. An adaptive large neighborhood search algorithm with three kinds of destroy operators and repair operators is developed to approximately solve the problem. Experiments based on lockage scheduling show that our algorithm found exact optimal solutions in most cases of no more than 30 items and outperformed typical genetic algorithm, simulated annealing, and two kinds of variable neighborhood descent algorithms in all cases of no more than 50 items.

With the continuous development of computer and network technology, the large‐scale and clustered operations of drones have gradually become a reality. How to realize the reasonable allocation of UAV cluster combat tasks and realize the intelligent optimization control of UAV cluster is one of the most challenging difficulties in UAV cluster combat. Solving the task allocation problem and finding the optimal solution have been proven to be an NP‐hard problem. This paper proposes a CSA‐based approach to simultaneously optimize four objectives in multi‐UAV task allocation, i.e., maximizing the number of successfully allocated tasks, maximizing the benefits of executing tasks, minimizing resource costs, and minimizing time costs. Experimental results show that, compared with the genetic algorithm, the proposed method has better performance on solving the UAV task allocation problem with multiple objectives.

With the rapid development of science and technology, unmanned technology has been widely used in many fields. One of the most important applications is in the field of civil and military UAVs. In the field of military UAVs (unmanned aerial vehicles), UAVs usually have to complete a series of tasks. In this series of tasks, there are often some key tasks. Key tasks play an important role, which is highly related to the feasibility of the whole action or task; mission failure sometimes causes incalculable damage. When assigning tasks to UAVs, it is necessary to ensure the accurate implementation of key tasks, so as to ensure the orderly implementation of the overall task. This paper not only successfully solved the previous problems but also comprehensively considered the minimization of resource consumption and the maximization of task revenue in the process of UAV task allocation. On the basis of considering the key system, considering the constraints and multiobjective problems in the UAV task allocation process, the violence allocation algorithm, constraint optimization evolutionary algorithm, PSO algorithm, and greedy algorithm combined with a constraint evolutionary algorithm are improved and optimized; it has been proven that they can solve the above difficulties. At the same time, several comparison experiments have been carried out; the performance and conclusion of the above four algorithms in the “limited” UAV task allocation scheme are analyzed in the experimental part.

A novel strength Pareto gravitational search algorithm (SPGSA) is proposed to solve multi-objective optimization problems. This SPGSA algorithm utilizes the strength Pareto concept to assign the fitness values for agents and uses a fine-grained elitism selection mechanism to keep the population diversity. Furthermore, the recombination operators are modeled in this approach to decrease the possibility of trapping in local optima. Experiments are conducted on a series of benchmark problems that are characterized by difficulties in local optimality, nonuniformity, and nonconvexity. The results show that the proposed SPGSA algorithm performs better in comparison with other related works. On the other hand, the effectiveness of two subtle means added to the GSA are verified, i.e. the fine-grained elitism selection and the use of SBX and PMO operators. Simulation results show that these measures not only improve the convergence ability of original GSA, but also preserve the population diversity adequately, which enables the SPGSA algorithm to have an excellent ability that keeps a desirable balance between the exploitation and exploration so as to accelerate the convergence speed to the true Pareto-optimal front.