In many practical manufacturing systems, transportation equipment such as automated guided vehicles (AGVs) is widely adopted to transfer jobs and realize the collaboration of different machines, but is often ignored in current researches. In this article, we address the flexible job shop scheduling problem with finite transportation resources (FJSP-Ts). Considering the difficulties caused by the introduction of transportation and the NP-hard nature, the evolutionary algorithm (EA) is adopted as a solution approach. To this end, a learning-based multipopulation evolutionary optimization (LMEO) is proposed to deal with the FJSP-T. First, the multipopulation strategy is introduced and a cooperation-based initialization is designed by combining several heuristics to guarantee the quality and diversity of the initial population. Second, a reinforcement learning (RL)-based mating selection is proposed to realize the cooperation of different subpopulations by selecting appropriate individuals for evolutionary search. Then, a specific local search inspired by the problem properties is designed to enhance the exploitation capability of the LMEO. Moreover, a statistical learning-based replacement is designed to maintain the quality and diversity of the population. Extensive experiments are conducted to test the performances of the LMEO. The statistical comparison shows that the LMEO is superior to the state-of-the-art algorithms in solving the FJSP-T in terms of solution quality and robustness.

With the development of intelligent transportation technologies, on-demand food delivery has become a prevalent logistics service that brings convenience to our daily life. However, problems such as inappropriate assignment and delayed delivery still exist in current food delivery platforms. To address these issues, this paper presents a Crowdsourced Recommending-and-Grabbing (CRG) system for order allocation in on-demand food delivery service. Unlike other traditional logistics systems, the CRG system considers the preferences of crowdsourced riders and allocates orders by recommending them to suitable crowdsourced riders. Riders can grab their favorite orders from personalized order lists. To optimize the experience of both customers and crowdsourced riders, a hierarchical solution framework composed of prediction and optimization is proposed to generate satisfactory order allocation schemes. First at the prediction part, we predict the preferences of crowdsourced riders by establishing a predictive model using eXtreme Gradient Boosting (XGBoost). Then at the optimization phase, an allocating-and-sequencing algorithm is designed to appropriately expose orders to crowdsourced riders and sequence the order list for each rider. Extensive experiments are conducted on real-world datasets and the comparative results demonstrate the effectiveness of the proposed methods, suggesting that the presented CRG system is promising to facilitate the on-demand food delivery effectively and efficiently.

On-demand food delivery (OFD) service is a representative scenario of last-mile logistics. It has gained a fast-growing market but also encounters many challenges, e.g., high dynamism, large-scale complexity, and immediacy requirement. To solve the OFD problem, a predictive-reactive optimization framework with feedback-based knowledge distillation is presented by organically combining deep learning technology and operational research method. In the prediction phase, a deep learning model is designed to predict future information which can reflect the delivery efficiency of dispatching results. To improve model performance, a feedback-based knowledge distillation is proposed which balances the diversity and effectiveness of the ensembled models by adaptively controlling learning weights. In the optimization phase, to avoid myopic decisions and obtain high-quality solutions for long-term objectives, a greedy heuristic with a multi-stage decision-making strategy is designed by employing the predicted future information to assist in making decisions. Extensive experiments are conducted on real-world datasets to test the performance of the proposed model and heuristic. Besides, the simulation results illustrate the superiority of the proposed framework for solving the OFD problem in both delivery efficiency and customer satisfaction.

Online food delivery (OFD) service has developed rapidly due to its great convenience for customers, the enormous markets for restaurants and the abundant job openings for riders. However, OFD platforms are encountering enormous challenges, such as massive demand, inevitable uncertainty and short delivery time. This article addresses an OFD problem with stochastic food preparation time. It is a complex NP-hard problem with uncertainty, large search space, strongly coupled subproblems, and high timeliness requirements. To solve the problem, we design an iterated greedy algorithm with a decomposition-based strategy. Concretely speaking, to cope with the large search space due to massive demands, a filtration mechanism is designed by preliminarily selecting suitable riders. To reduce the risk affected by the uncertainty, we introduce a risk-measuring criterion into the objective function and employ a scenario-sampling method. For timeliness requirements caused by short delivery time, we design two time-saving strategies via mathematical analysis, i.e., an adaptive selection mechanism to choose the method with less computational effort and a fast evaluation mechanism based on the small-scale sampling and machine learning model to speed up evaluation. We also prove an upper bound of the stochastic time cost under risk measurement as one of the baseline features to improve the prediction accuracy. The experiments on real-world data sets demonstrate the effectiveness of the proposed algorithm.

As one representative of the emerging on-demand transport services, the on-demand food delivery (OFD) has penetrated into daily life. Due to its intrinsic complexities, the OFD has attracted the interest of a growing number of logistics researchers. This paper aims at optimizing the OFD process and addresses an OFD problem (OFDP). To overcome the dynamic and large-scale complexity, we abstract the OFDP into a static generalized assignment problem with a rolling horizon strategy. To meet the demand on high service quality and limited computation time, we propose an offline-optimization for online-operation framework based on imitation learning. Under this framework, an imitation learning-enhanced iterated matching algorithm (ILIMA) is proposed, which consists of three basic components: an iterated matching heuristic (IMH) to fast generate solutions, an expert to provide expertise, and a machine learning (ML) model to assist the decision-making process in IMH by mimicking the expert. In the offline-optimization phase, the ML model mines knowledge from the high-quality solutions optimized by the expert; in the online-operation phase, the IMH embedded with the well-trained ML model is deployed online to make decisions in a real OFD scenario. Offline simulation experiments are carried out on real historical data, which validate the superiority of ILIMA compared with existing methods. Moreover, rigorous online A/B tests are conducted on the scheduling system of Meituan, which demonstrates the practical value of ILIMA to improve customer satisfaction and delivery efficiency.

In recent years, the online food ordering (OFO) platforms have arose fast and brought huge convenience to people in daily life. Under the scenario of a realistic OFO platform, this paper addresses an online meal delivery problem (OMDP). To reduce the search space, the OMDP is decomposed into two sub-problems, i.e., the pickup and delivery problem and the order dispatching problem. To solve each sub-problem effectively, a hybrid differential evolution algorithm is proposed, which is fused by the DE-based phase to plan routes and the heuristic-based phase to determine order dispatching schemes. In the DE-based routing phase, a heuristic considering the urgency of orders is designed to generate the initial population with certain quality. Besides, a mutation operator is developed to enhance the exploration and a crossover operator embedded with local search is designed to enhance the exploitation. In the heuristic-based dispatching phase, a regret heuristic is presented to produce good dispatching solutions by introducing the influences between delivery persons. Numerical tests have been carried out and computational results demonstrate the effectiveness of the proposed algorithm.

Considering the uncertainty in the real world online food delivery applications, this paper addresses an online order dispatching problem with fuzzy preparation times (FOODP). According to the characteristics of the problem, the FOODP is decomposed into two sub-problems, an order assignment problem and a fuzzy traveling salesman problem with pickup and delivery. To deal with the two sub-problems efficiently, a two-stage algorithm is proposed by reasonably fusing a modified greedy search (mGS) and the fruit fly optimization algorithm (FOA). In the mGS phase, agreement index is employed in the multi-stage decision to search for robust optimal solutions. In the FOA-based search phase, a modified heuristic is used to generate the initial route. To enhance the exploration of the algorithm, an olfactory search is designed by the cooperation of several problem-specific search operators. Moreover, a specific local intensification is employed to further improve the performance of solutions. Numerical tests and statistical analysis demonstrate the effectiveness and efficiency of the proposed algorithm.

BACKGROUND: Stem cells and stem cell lines are widely used in biomedical research. The Cell Ontology (CL) and Cell Line Ontology (CLO) are two community-based OBO Foundry ontologies in the domains of in vivo cells and in vitro cell line cells, respectively. RESULTS: To support standardized stem cell investigations, we have developed an Ontology for Stem Cell Investigations (OSCI). OSCI imports stem cell and cell line terms from CL and CLO, and investigation-related terms from existing ontologies. A novel focus of OSCI is its application in representing metadata types associated with various stem cell investigations. We also applied OSCI to systematically categorize experimental variables in an induced pluripotent stem cell line cell study related to bipolar disorder. In addition, we used a semi-automated literature mining approach to identify over 200 stem cell gene markers. The relations between these genes and stem cells are modeled and represented in OSCI. CONCLUSIONS: OSCI standardizes stem cells found in vivo and in vitro and in various stem cell investigation processes and entities. The presented use cases demonstrate the utility of OSCI in iPSC studies and literature mining related to bipolar disorder.

BACKGROUND: Different human responses to the same vaccine were frequently observed. For example, independent studies identified overlapping but different transcriptomic gene expression profiles in Yellow Fever vaccine 17D (YF-17D) immunized human subjects. Different experimental and analysis conditions were likely contributed to the observed differences. To investigate this issue, we developed a Vaccine Investigation Ontology (VIO), and applied VIO to classify the different variables and relations among these variables systematically. We then evaluated whether the ontological VIO modeling and VIO-based statistical analysis would contribute to the enhanced vaccine investigation studies and a better understanding of vaccine response mechanisms. RESULTS: Our VIO modeling identified many variables related to data processing and analysis such as normalization method, cut-off criteria, software settings including software version. The datasets from two previous studies on human responses to YF-17D vaccine, reported by Gaucher et al. (2008) and Querec et al. (2009), were re-analyzed. We first applied the same LIMMA statistical method to re-analyze the Gaucher data set and identified a big difference in terms of significantly differentiated gene lists compared to the original study. The different results were likely due to the LIMMA version and software package differences. Our second study re-analyzed both Gaucher and Querec data sets but with the same data processing and analysis pipeline. Significant differences in differential gene lists were also identified. In both studies, we found that Gene Ontology (GO) enrichment results had more overlapping than the gene lists and enriched pathway lists. The visualization of the identified GO hierarchical structures among the enriched GO terms and their associated ancestor terms using GOfox allowed us to find more associations among enriched but often different GO terms, demonstrating the usage of GO hierarchical relations enhance data analysis. CONCLUSIONS: The ontology-based analysis framework supports standardized representation, integration, and analysis of heterogeneous data of host responses to vaccines. Our study also showed that differences in specific variables might explain different results drawn from similar studies.

Biomedical ontologies are often very large and complex. Only a subset of the ontology may be needed for a specified application or community. For ontology end users, it is desirable to have community-based labels rather than the labels generated by ontology developers. Ontodog is a web-based system that can generate an ontology subset based on Excel input, and support generation of an ontology community view, which is defined as the whole or a subset of the source ontology with user-specified annotations including user-preferred labels. Ontodog allows users to easily generate community views with minimal ontology knowledge and no programming skills or installation required. Currently >100 ontologies including all OBO Foundry ontologies are available to generate the views based on user needs. We demonstrate the application of Ontodog for the generation of community views using the Ontology for Biomedical Investigations as the source ontology.

The wide-scale development of ontologies in the bioinformatics domain facilitates their use in the creation of scientific workflows. To speed up the design of workflows, a Service Suggestion Engine is interfaced to the Galaxy Tool Integration and Workflow Platform. This enables users to ask for suggestions (e.g., what operation should go next) while designing workflows with the Galaxy user interface. The Service Suggest Engine utilizes semantic annotations to suggest appropriate Web service operations to plug into the workflow under design. The enriched Ontology for Biomedical Investigation (OBI) is used as a target for the annotations. The effectiveness of the suggestions provided is evaluated against a consensus of domain experts.

MOTIVATION: Describing biological sample variables with ontologies is complex due to the cross-domain nature of experiments. Ontologies provide annotation solutions; however, for cross-domain investigations, multiple ontologies are needed to represent the data. These are subject to rapid change, are often not interoperable and present complexities that are a barrier to biological resource users. RESULTS: We present the Experimental Factor Ontology, designed to meet cross-domain, application focused use cases for gene expression data. We describe our methodology and open source tools used to create the ontology. These include tools for creating ontology mappings, ontology views, detecting ontology changes and using ontologies in interfaces to enhance querying. The application of reference ontologies to data is a key problem, and this work presents guidelines on how community ontologies can be presented in an application ontology in a data-driven way. AVAILABILITY: http://www.ebi.ac.uk/efo.