Task Decomposition Method Research Articles

Two-stage task allocation for multiple construction robots using an improved genetic algorithm

The construction industry is facing serious challenges, such as labor shortage and safety hazards. Construction robots can perform dangerous and repetitive tasks instead of workers. However, complex construction tasks require multi-robot collaboration, which needs efficient task allocation. This paper describes a two-stage task allocation method using an improved Genetic Algorithm (GA) for multiple construction robots. First, a construction task decomposition and grouping method under construction constraints is established, using a steel structure as example. Second, a Multi-Robot Construction Task Allocation (MRCTA) model is developed by modeling construction robots and tasks. Third, the MRCTA problem is divided into two stages, i.e., determining the optimal construction robot group for each task and the best task sequence for each robot group. Finally, a simulated experiment was designed and conducted. The results show that the method can efficiently find optimal solutions and the improved GA can enhance the diversity to achieve well-adapted solutions.

Faster MIL-based Subgoal Identification for Reinforcement Learning by Tuning Fewer Hyperparameters

Various methods have been proposed in the literature for identifying subgoals in discrete reinforcement learning (RL) tasks. Once subgoals are discovered, task decomposition methods can be employed to improve the learning performance of agents. In this study, we classify prominent subgoal identification methods for discrete RL tasks in the literature into the following three categories: graph-based, statistics-based, and multi-instance learning (MIL)-based. As contributions, firstly, we introduce a new MIL-based subgoal identification algorithm called EMDD-RL and experimentally compare it with a previous MIL-based method. The previous approach adapts MIL’s Diverse Density (DD) algorithm, whereas our method considers Expected-Maximization Diverse Density (EMDD). The advantage of EMDD over DD is that it can yield more accurate results with less computation demand thanks to the expectation-maximization algorithm. EMDD-RL modifies some of the algorithmic steps of EMDD to identify subgoals in discrete RL problems. Secondly, we evaluate the methods in several RL tasks for the hyperparameter tuning overhead they incur. Thirdly, we propose a new RL problem called key-room and compare the methods for their subgoal identification performances in this new task. Experiment results show that MIL-based subgoal identification methods could be preferred to the algorithms of the other two categories in practice.

NOx emissions prediction for MSWI process based on dynamic modular neural network

Accurate prediction of nitrogen oxides (NOx) is crucial for improving the efficiency of denitrification systems in municipal solid waste incineration (MSWI) process. Due to the change of feed composition and operation mode, it is difficult to predict NOx emissions with complex dynamics. For this reason, a dynamic modular neural network (DMNN) is proposed for NOx emissions prediction in MSWI process. First, a principal component analysis (PCA)-based dynamic task decomposition method is proposed, and then the original task with time varying characteristic is divided into several sub-tasks for effectual handing. Next, an adaptive long short-term memory (ALSTM) network is designed driven by the corresponding sub-task. Then, the nonlinearity between dominant variables and NOx value is learned to guarantee the prediction accuracy. Finally, the merits of proposed DMNN are confirmed on a benchmark and real industrial data of a MSWI process. The experimental results further demonstrate the superiority and potential of DMNN for industrial applications.

Cooperative modular reinforcement learning for large discrete action space problem

Deep reinforcement learning (DRL) has achieved remarkable results on high-dimension state tasks. However, it suffers in hard convergence and low sample efficiency when solving large discrete action space problems. To meet these challenges, we develop a cooperative modular reinforcement learning (CMRL) method to distributedly solve the problems with a large discrete action space. A general yet effective task decomposition method is proposed to decompose the complex decision task in a large action space into multiple decision sub-tasks in small action subsets, using a rule-based action division method. The CMRL method consisting of multiple Critic networks is proposed to settle the multiple sub-tasks, where each Critic network learns a decomposed value function to obtain the local optimal action in a sub-task. The global optimal action is cooperatively chosen by all local optimal actions. Moreover, we propose a new parallel training mechanism, which trains multiple Critic networks with different models and multi-data in parallel. Mathematical properties are proposed to analyze the rationality and superiority of CMRL. Four different simulation experiments are conducted to verify the generality and effectiveness of CMRL for large action space problems. The results show that CMRL has superior performance on training efficiency compared with classical and latest DRL methods while maintaining the accuracy of the solution.

Structure Learning-Based Task Decomposition for Reinforcement Learning in Non-stationary Environments

Reinforcement learning (RL) agents empowered by deep neural networks have been considered a feasible solution to automate control functions in a cyber-physical system. In this work, we consider an RL-based agent and address the issue of learning via continual interaction with a time-varying dynamic system modeled as a non-stationary Markov decision process (MDP). We view such a non-stationary MDP as a time series of conventional MDPs that can be parameterized by hidden variables. To infer the hidden parameters, we present a task decomposition method that exploits CycleGAN-based structure learning. This method enables the separation of time-variant tasks from a non-stationary MDP, establishing the task decomposition embedding specific to time-varying information. To mitigate the adverse effect due to inherent noises of task embedding, we also leverage continual learning on sequential tasks by adapting the orthogonal gradient descent scheme with a sliding window. Through various experiments, we demonstrate that our approach renders the RL agent adaptable to time-varying dynamic environment conditions, outperforming other methods including state-of-the-art non-stationary MDP algorithms.

Understanding skill decay and skill maintenance in first responders

To carry out their roles, first responders need to have appropriate skills. It is important to identify and train in key skills appropriate for the tasks that first responders will perform. Once these skills have been acquired, they need to be maintained otherwise they will decay to the point where performance of the skill is no longer acceptable. This means that emergency services organisations need programs that appropriately maintain the skills of their volunteers and employees. To deliver cost-effective training, these organisations need a good understanding of what key skills are required for tasks that are regularly performed, why and how quickly these skills decay and how these skills can be maintained. To help emergency services organisations better understand these important concepts, this paper reviews relevant literature on skill decay and skill maintenance. Task decomposition methods and training needs analysis are introduced to assist organisations determine what key skills they require. This provides information to make sound, evidence-based decisions about recurrent training programs that can maintain the skills required by first responders and retain efficacy in the organisation.

Task Decomposition Based on Cloud Manufacturing Platform

As a new service-oriented manufacturing paradigm, cloud manufacturing (CMfg) realizes the optimal allocation of resources in the product manufacturing process through the network. Task decomposition is a key problem of the CMfg system for resource scheduling. A high-quality task decomposition method can shorten product development time, reduce costs for resource service providers, and provide technical support for the application of CMfg. However, a cloud manufacturing system has to manage the allocation the correct amount of manufacturing resources, complex production processes, and highly dynamic production environments. At the same time, the tasks issued by service demanders are usually asymmetric and tightly coupled. We solve the complex task decomposition problem by using the traditional methods, that are hard to complete in CMfg. To overcome the shortcomings of CMfg, this paper proposed a task decomposition method based on the cloud platform. For achieving modular production, this approach creatively divides the product production process into four stages: design, manufacturing, transportation, and maintenance. Then a hybrid method, which combines with depth-first search algorithm, fast modular optimization algorithm, and artificial bee colony algorithm, is introduced. The method can obtain a multi-stage task optimization decomposition plan in CMfg. Simulation results demonstrate the proposed method can achieve complex task optimization decomposition in a CMfg environment.

Task decomposition and distribution method for heterogeneous multi-robotic systems

This article presents a method for decomposition and distribution of tasks in a multi-robotic system (MRS). The proposed implementation of the method of elementary operations distribution is based on maximizing the expected efficiency of each agent and collective decision making (CDM). To compare the effectiveness of various methods, computational experiments were carried out. According to the results, the implementation of the CDM mechanisms allows reducing the time for performing a global task by MRS agents by more than 10%. For the initial conditions described in the work, the use of the proposed method for maximizing the individual efficiency of an agent in combination with the CDM algorithm allows reducing the average execution time of a global operation by 21% in comparison with the method of minimizing the individual execution time without CDM.

High-end equipment development task decomposition and scheme selection method

Decomposition of tasks and selection of optimal schemes are key procedures in high-end equipment development processes. However, such procedures are highly innovative, technology intensive, interdisciplinary, and multi-party engineering projects, making the decomposition and scheme selection more difficult and complicated than that in the development of ordinary equipment. In this study, we consider three factors, namely, functional structure, task granularity, and task feasibility in task decomposition of high-end equipment development. Based on the principles of systems engineering, a method of task decomposition is proposed. As for decomposition scheme selection, a method based on the superiority and inferiority ranking (SIR) method mixed information and multiple attribute decision making is proposed by considering attributes of scheme feasibility, uncertainty risk and task integration complexity. To verify the proposed method, development of a military electric vehicle is used as an example to demonstrate the calculation process.

Development and Validation of a Descriptive Cognitive Model for Predicting Usability Issues in a Low-Code Development Platform.

The aim of the study was the development and evaluation of a Descriptive Cognitive Model (DCM) for the identification of three types of usability issues in a low-code development platform (LCDP). LCDPs raise the level of abstraction of software development by freeing end-users from implementation details. An effective LCDP requires an understanding of how its users conceptualize programming. It is necessary to identify the gap between the LCDP end-users' conceptualization of programming and the actions required by the platform. It is also relevant to evaluate how the conceptualization of the programming tasks varies according to the end-users' skills. DCMs are widely used in the description and analysis of the interaction between users and systems. We propose a DCM which we called PRECOG that combines task decomposition methods with knowledge-based descriptions and criticality analysis. This DCM was validated using empirical techniques to provide the best insight regarding the users' interaction performance. Twenty programmers (10 experts, 10 novices) were observed using an LCDP and their interactions were analyzed according to our DCM. The DCM correctly identified several problems felt by first-time platform users. The patterns of issues observed were qualitatively different between groups. Experts mainly faced interaction-related problems, while novices faced problems attributable to a lack of programming skills. By applying the proposed DCM we were able to predict three types of interaction problems felt by first-time users of the LCDP. The method is applicable when it is relevant to identify possible interaction problems, resulting from the users' background knowledge being insufficient to guarantee a successful completion of the task at hand.

Task Decomposition for MPC: A Computationally Efficient Approach for Linear Time-Varying Systems

Abstract A Task Decomposition method for iterative learning Model Predictive Control (TDMPC) for linear time-varying systems is presented. We consider the availability of state-input trajectories which solve an original task τ1, and design a feasible MPC policy for a new task, τ2, using stored data from τ1. Our approach applies to tasks τ2 which are composed of subtasks contained in τ1. In this paper we formally define the task decomposition problem, and provide a feasibility proof for the resulting policy. The proposed algorithm reduces the computational burden for linear time-varying systems with piecewise convex constraints. Simulation results demonstrate the improved efficiency of the proposed method on a robotic path-planning task.

A Novel k-Means Clustering Based Task Decomposition Method for Distributed Vector-Based CA Models

More and more vector-based cellular automata (VCA) models have been built to leverage parallel computing to model rapidly changing cities and urban regions. During parallel simulation, common task decomposition methods based on space partitioning, e.g. grid partitioning (GRID) and recursive binary space partitioning (BSP), do not work well given the heterogeneity of VCA parcel tasks. In this paper, to solve this problem, we propose a novel task decomposition method for distributed VCA models based on k-means clustering, named KCP. Firstly, the polygon dataset is converted into points based on centroids, which combines the size of two parcels and the outer distance. A low-cost recursive quad-partition is then applied to decide the initial cluster centers based on parcel density. Finally, neighbor parcels can be allocated into the same subdivision through k-means clustering. As a result, the proposed KCP method takes both the number of tasks and computing complexity into consideration to achieve a well-balanced local workload. A typical urban VCA growth model was designed to evaluate the proposed KCP method with traditional spatial partitioning methods, i.e. GRID and BSP. KCP had the shortest total simulation time when compared with GRID and BSP. During experimental urban growth simulations, the time spent on a single iteration was reduced by 15% with the BSP and by 25% with the GRID method. The total simulation time with a 120 m neighborhood buffer size was reduced by more than one hour to around three minutes with 32 cores.

Multisensors Cooperative Detection Task Scheduling Algorithm Based on Hybrid Task Decomposition and MBPSO

A multisensor scheduling algorithm based on the hybrid task decomposition and modified binary particle swarm optimization (MBPSO) is proposed. Firstly, aiming at the complex relationship between sensor resources and tasks, a hybrid task decomposition method is presented, and the resource scheduling problem is decomposed into subtasks; then the sensor resource scheduling problem is changed into the match problem of sensors and subtasks. Secondly, the resource match optimization model based on the sensor resources and tasks is established, which considers several factors, such as the target priority, detecting benefit, handover times, and resource load. Finally, MBPSO algorithm is proposed to solve the match optimization model effectively, which is based on the improved updating means of particle’s velocity and position through the doubt factor and modified Sigmoid function. The experimental results show that the proposed algorithm is better in terms of convergence velocity, searching capability, solution accuracy, and efficiency.

Role-based approaches for operational tasks modeling and flexible decomposition

In order to realize the fact that operational tasks can be decomposed flexibly with new tasks random adding, role-based approaches for operational tasks model and the flexibility decomposition are proposed. Firstly, aiming at the shortcomings of previous task decomposition methods, operational role concepts are introduced to design a computing framework of task flexibility decomposition. Within the framework, task structure matrixes are constructed for tasks decomposition with the mapping relations of information interaction and action process among tasks. On basis of this, via giving operational decomposition granularity, a flexibility approach of operational task decomposition is put forward with three parts about task decomposition granularity adjustment, tasks module partition with different operational role patterns and task decomposition scheme evaluation. Finally, as an application case shown, it is validated that the proposed approaches for tasks modeling and decomposition are available. By comparing with previous decomposition approaches, the proposed approach, within three types of operational roles, has strong environmental adaptability and scheme diversity, which can restrain calculation complexity increment for realizing task flexible decomposition effectively.

Task Decomposition and Grouping for Customer Collaboration in Product Development

AbstractDecomposing product development project into various tasks and grouping them are important activities in product development. Many scholars devoted their efforts to solving this problem and proposed some useful methods. However, the research work of task decomposition and grouping for customer collaboration in product development is still lacking. Therefore, this study first decomposes product development tasks and analyzes its executability. Then, by using an integrated numerical design structure matrix and adaptive genetic algorithm (AGA) approach, tasks are divided into different groups, tasks in the same group have high correlation degree, and tasks in the different groups have low correlation degree. To illustrate the process of task decomposition and grouping methods proposed in this paper, a mobile phone structural development case is applied. Moreover, standard generic algorithm (SGA) and particle swarm optimization (PSO) are used to compare with AGA to verify the effectiveness of AGA.

Task Decomposition Method of R&D Project Based on Product Structure Tree

To improve the accuracy and efficiency of task decomposition of R&D (Research and Development) projects, a new decomposition method based on product structure tree was proposed. The relevance between product structure tree and work breakdown structure (WBS) was analyzed through their natural quality and association relations. The mapping rules between product components and project tasks were defined. Mathematical models about the mapping from components and parts to tasks and combining tasks were established. Based on the above mechanism and models, the automatic mapping transformation from product components to project tasks was achieved. A prototype system was developed based on the task decomposition method, and the system had been applied in R&D projects of a diesel engine factory, which verified its effectiveness and feasibility.

Extension of automated melanoma screening for non-melanocytic skin lesions

In this paper, we present an automated melanoma screening system that supports not only melanocytic skin lesions (MSLs) but also non-melanocytic skin lesions (NoMSLs). Melanoma is known as the most fatal skin cancer. Therefore, early detection is highly desired. However, melanoma diagnosis is not easy even for expert dermatologists. Several researchers have developed automated methods for melanoma detection but they mostly focused only on MSLs while NoMSLs have been almost neglected. To expand the scope to NoMSLs, we developed two melanoma classification models, namely the single-shot model and the double-shot model. The single-shot model differentiates melanomas from all the other skin lesions including NoMSLs. The double-shot model divides the task into two subtasks. Firstly, it differentiates MSLs from NoMSLs and then differentiates melanomas from the other MSLs. The single-shot achieved a sensitivity (SE) of 92.9% and a specificity (SP) of 83.9%, while the double-shot achieved an SE of 97.6% and an SP of 87.7% with ten image features. Through a detailed data analysis, we confirmed that the double-shot can exploit effective image features more than the single-shot as a result of its task decomposition method.

A Group Cooperative Decision Support System Based on Extended Contract Net

With the prevalence of the Web, most decision-makers are likely to use the Web to support their decision-making. Furthermore, many decision-makers begin to seek cooperators on the Internet to help them to solve complex decision problem. In the paper, we propose a research framework for Group Cooperative Decision Support System (GCDSS) and determine three layers such as task layer, cooperator layer and resource layer to characterize it and demonstrate how electronic market could facilitate cooperation in GCDSS. Then we present the negotiation model of GCDSS and propose a task decomposition method based on which one can decompose a complex problem into a formal task AND/OR tree. We also propose a task assignment method based on extended Contract Net to perform task allocation in GCDSS and describe the group cooperative decision-making process in Knowledge Query and Manipulation Language (KQML). An optimal policy is designed to find the optimal solution of task AND/OR tree and cooperative problem solving process is illustrated with an example for education development plan in China. Finally, we conclude with a brief overview of research contribution, along with some proposals for further research.

Multi-core real-time scheduling for generalized parallel task models

Multi-core processors offer a significant performance increase over single-core processors. They have the potential to enable computation-intensive real-time applications with stringent timing constraints that cannot be met on traditional single-core processors. However, most results in traditional multiprocessor real-time scheduling are limited to sequential programming models and ignore intra-task parallelism. In this paper, we address the problem of scheduling periodic parallel tasks with implicit deadlines on multi-core processors. We first consider a synchronous task model where each task consists of segments, each segment having an arbitrary number of parallel threads that synchronize at the end of the segment. We propose a new task decomposition method that decomposes each parallel task into a set of sequential tasks. We prove that our task decomposition achieves a resource augmentation bound of 4 and 5 when the decomposed tasks are scheduled using global EDF and partitioned deadline monotonic scheduling, respectively. Finally, we extend our analysis to a directed acyclic graph (DAG) task model where each node in the DAG has a unit execution requirement. We show how these tasks can be converted into synchronous tasks such that the same decomposition can be applied and the same augmentation bounds hold. Simulations based on synthetic workload demonstrate that the derived resource augmentation bounds are safe and sufficient.

HPCCD: Hybrid Parallel Continuous Collision Detection using CPUs and GPUs

AbstractWe present a novel, hybrid parallel continuous collision detection (HPCCD) method that exploits the availability of multi‐core CPU and GPU architectures. HPCCD is based on a bounding volume hierarchy (BVH) and selectively performs lazy reconstructions. Our method works with a wide variety of deforming models and supports self‐collision detection. HPCCD takes advantage of hybrid multi‐core architectures – using the general‐purpose CPUs to perform the BVH traversal and culling while GPUs are used to perform elementary tests that reduce to solving cubic equations. We propose a novel task decomposition method that leads to a lock‐free parallel algorithm in the main loop of our BVH‐based collision detection to create a highly scalable algorithm. By exploiting the availability of hybrid, multi‐core CPU and GPU architectures, our proposed method achieves more than an order of magnitude improvement in performance using four CPU‐cores and two GPUs, compared to using a single CPU‐core. This improvement results in an interactive performance, up to 148 fps, for various deforming benchmarks consisting of tens or hundreds of thousand triangles.