Performance Evaluation Indicator (PEI): A new paradigm to evaluate the competence of machine learning classifiers in predicting rockmass conditions

To illustrate an unprejudiced comparison among machine learning classifiers established on proprietary databases, and to guarantee the validity and robustness of these classifiers, a Performance Evaluation Indicator (PEI) and the corresponding failure criterion are proposed in this study. Three types of machine learning classifiers, including the strictly binary classifier, the normal multiclass classifier and the misclassification cost-sensitive classifier, are trained on four datasets recorded from a water drainage TBM project. The results indicate that: (1) the PEI successfully compares the competence of classifiers under different scenarios by isolating the effects of different overlapping-degree of rockmass classes, and (2) the cost-sensitive algorithm is warranted to classify rockmasses when the ratio of inter-class classes is more than 8:1. The contributions of this research are to fill the gap in performance evaluations of a classifier for imbalanced training data, and to identify the best situation to apply this classifier.     

A Building Information Model enabled Multiple Traveling Salesman Problem for building interior patrols

During building emergencies, an effective and visible primary search plan enhances situation awareness and enables a more efficient rescue mission. The aim of the primary search during an emergency is the rapid screening of every space in the building to identify locations of victims and their conditions. Afterwards, first responders can plan for the rescue of those victims. To provide a timely draw up of interior patrol routes and assign rescue teams to conduct the primary search, this study formulates the problem as a multiple traveling salesman problem (M-TSP) where the comprehensive building interior network is given by the building information models (BIMs), while the total traveling costs (lengths) of every rescue team is minimized. To meet the requirement of real-time patrol routes optimization, we employed the branch-and-price algorithm for the enhancement of computation efficiency. In addition, a heuristic method was introduced to provide timely solutions for large-scale networks. A case study is conducted for a single-floor convention center. We utilized BIM to construct a network of nodes and arcs where the decision model requires as input, and the branch-and-price algorithm finds the optimal patrol. The resulting patrol routes can be visualized and serve as guide for rescue teams to conduct the primary search. The integrated approach proposed in this study is practical and can expedite search and rescue missions.     

A systematic data-mining-based methodology for product family design and product configuration

Product family design and product configuration based on data mining technology is identified as an intelligent and automated means to improve the efficiency of product development. However, few of previous literatures have proposed systematic product family design method based on data mining technology. To make up for this deficiency, this research put forward a systematic data-mining-based method for product family design and product configuration. First, the customer requirement information and product engineering information in the historical order are formatted into structural data. Second, principal component analysis is performed on historical orders to extract the customers' differentiated needs. Third, association rule algorithm is introduced to mine the rules between differentiated needs and module instances in the historical orders, thus obtained the configuration knowledge between customer needs and product engineer. Forth, the mined rules are used to construct association rule-based classifier (CBA) that is employed to sort out the best product configuration schemes as popular product variants. Fifth, sequence alignment technique is employed to identify modules for popular product variants, so that the module instances are divided into optional, common and special module, respectively, thereby the product platform is generated based on common modules. Finally, according to new customer needs, the CBA classifier is used to recommend the best configuration schemes, and then popular product variants are configured based on the product platform. The feasibility of the proposed method is demonstrated by the product family design example of desktop computer hosts.     

Visitor-artwork network analysis using object detection with image-retrieval technique

Recent museum exhibitions are becoming a means by which to satisfy visitor demands. In order to provide visitor-centric exhibitions, artwork must be analyzed based on the behavior of visitors, and not merely according to museum professionals' points of view. This study aims to analyze the relationship between museum visitors and artwork via a network analysis based on visitor behavior using object detection techniques. Cameras installed in a museum recorded visitors, and an object detector with a content-based image-retrieval technique tracked visitors from the videos. The durations spent with different artworks were measured, and the data was converted into a bipartite graph. The relationships between different artwork types were analyzed with a visitor-centered artwork network. Based on the visitors’ behavior, significant artworks were identified and the artwork network was compared to the arrangement of the museum. The tendency of edges in the artwork network was also examined considering visitors' preferences for artworks. The method used here makes it possible to collect quantitative data, with the results possibly used as a basis and for reference when analyzing artwork in a visitor-centered approach.     

A synthetical development approach for rehabilitation assistive smart product–service systems: A case study

Digital transformation (DT), the combination of information, computing, communication and connectivity technologies, which has triggered an effective upgrade of different aspects of market strategy, customer experience etc. Nowadays, rehabilitation assistive devices (RADs) are evolving to be more digital, intelligent and personalized. Digitalization and servitization have fostered to an emerging business model—the smart product–service system (Smart PSS). Therefore, DT of the RADs’ industry advocates not only the design of products and functions, the more important is the management of service processes and resource integration. With the increase in the elderly and disabled population, the requirement for RADs is becoming more urgent. However, research on Smart PSS for RAD is still limited. The rehabilitation assistive smart product–service systems (RASPSS) was introduced into the development of RADs based on the “Design and Management of DT” strategy through the service design of assistive devices and user requirements analysis. Further, an integrated design of RAD and Smart PSS has been created, a development method of RASPSS proposed, the theoretical model of the Smart PSS based on RADs built. To specify the service framework, this case study discusses the development of a home rehabilitation assistive system for femoral stem fracture patients. This paper evaluates the usability of the system, the results of which prove usability and effectiveness of the RASPSS development method. The RASPSS development model is designed to meet needs of stakeholders, improve the user rehabilitation experience, promote the service innovation of Smart PSS, bring certain market benefits of rehabilitation aids and create social value.     

Smart dispatching and optimal elevator group control through real-time occupancy-aware deep learning of usage patterns

Passengers spend much time on elevator journeys in high-rise buildings every day, in which unnecessary stops caused by lack of cab capacity take up a certain proportion of the journey time. This study proposes real-time occupancy-aware smart dispatching to avoid pick-up failure by introducing the occupancy information that reflects elevator capacity into the optimization model, thus improving dispatching performance. Occupancy awareness is firstly implemented with deep learning-based object detection to provide estimated capacity. Traffic pattern recognition is implemented with time series analysis and fuzzy logic. Case-based reasoning is applied to recognize the current usage pattern and to deploy specific dispatching strategies. A prioritized A* search model is built to solve dispatching optimization with occupancy information. Discrete event simulation is conducted with Simio and MATLAB to validate the proposed dispatching model.     

Dynamic ensemble wind speed prediction model based on hybrid deep reinforcement learning

Prediction of wind speed can provide a reference for the reliable utilization of wind energy. This study focuses on 1-hour, 1-step ahead deterministic wind speed prediction with only wind speed as input. To consider the time-varying characteristics of wind speed series, a dynamic ensemble wind speed prediction model based on deep reinforcement learning is proposed. It includes ensemble learning, multi-objective optimization, and deep reinforcement learning to ensure effectiveness. In part A, deep echo state network enhanced by real-time wavelet packet decomposition is used to construct base models with different vanishing moments. The variety of vanishing moments naturally guarantees the diversity of base models. In part B, multi-objective optimization is adopted to determine the combination weights of base models. The bias and variance of ensemble model are synchronously minimized to improve generalization ability. In part C, the non-dominated solutions of combination weights are embedded into a deep reinforcement learning environment to achieve dynamic selection. By reasonably designing the reinforcement learning environment, it can dynamically select non-dominated solution in each prediction according to the time-varying characteristics of wind speed. Four actual wind speed series are used to validate the proposed dynamic ensemble model. The results show that: (a) The proposed dynamic ensemble model is competitive for wind speed prediction. It significantly outperforms five classic intelligent prediction models and six ensemble methods; (b) Every part of the proposed model is indispensable to improve the prediction accuracy.     

Hierarchical control for cornering stability of dual-motor RWD vehicles with electronic differential system using PSO optimized SOSMC method

This paper proposes a novel control scheme with a three-layer hierarchical structure to improve the cornering stability of the dual-motor rear-wheel drive (RWD) vehicles with the electronic differential system (EDS). The proposed hierarchical structure for the control system includes the observing layer, control layer, and actuation layer. In the observing layer, the driver model is designed to obtain the nominal steering angle, and the state observer is designed to obtain the yaw angle which cannot be easily measured. Then, particle swarm optimization (PSO) and second order sliding mode control (SOSMC) are employed in the control layer. The SOSMC part is used to design the control law to eliminate the chattering problem in the sliding mode algorithm, and the PSO part is used to obtain the optimal weights in the sliding mode surface to meet the minimum sideslip angle error and yaw rate error. The actuation layer allocates the corrected yaw moment by distributing the driving force to each independent driving wheel. Finally, the numerical tests are carried out under the double line change (DLC) maneuver. The results show that the proposed control system can effectively improve the cornering stability of the dual-motor RWD vehicles and reduce their motor power consumption.     

Artificial intelligence for BIM content management and delivery: Case study of association rule mining for construction detailing

The proliferation of Building Information Modeling (BIM) applications, in tandem with the extensive variation of building products, pose new demands on design and engineering firms to efficiently manage and reuse BIM content (i.e., data-rich parametric model objects and assembly details). Tasks such as classifying BIM objects, indexing them with meta-data (e.g., category), and searching digital libraries to load objects into models still plague practice with inefficient manual workflows. This research aims to improve the productivity of BIM content management and retrieval by developing an AI-backed BIM content recommender system. Using data from a case-study firm, this research extracted content from over 30,000 technical BIM views (e.g., plans, sections, details) in historical projects to build an unsupervised machine-learning prototype with association rule mining. This prototype explicated the strength of relationships among co-occurring BIM objects. Using this prototype as the backbone AI-engine in live BIM sessions, this research developed a context-aware recommender system that dynamically provides BIM users with a set of objects associable with their modeling context (e.g., type of view, existing objects in the model) and human–computer interactions (e.g., objects selected by the user). By mining association data from hundreds of historical projects, this development marks a departure from the existing prototypes that rely on explicit coding, recurring user input, or subjective ratings to recommend BIM content to users. The simulation and experimental implementation of this recommender system yielded high efficacy in predicting content needs and achieved significant savings in the time spent on conventional BIM workflows.     

Context-aware cognitive design assistant: Implementation and study of design rules recommendations

Design rules are an essential interface to facilitate the information exchange between designers and experts. Despite many innovations in Knowledge-Based Engineering and Knowledge Management, unstructured design rules documents are still widely used in the manufacturing industry. Due to the complexity of the design process, these documents often contain hundreds of design rules, applicable in varying design contexts. Searching for the right rules according to a design context is demanding in time and cognitive resources. In this paper, we propose a Context-Aware Cognitive Design Assistant (CACDA) to capture the design context and perform tasks such as the recommendation of design rules, the verification of design solutions, or the automation of design routines. Contrary to existing works in model quality testing, the CACDA uses a proactive approach of design rules application and helps designers to provide error-free designs on first attempt. In this paper, we present the design rules recommendation system of the CACDA, its capabilities and its implementation. Then, to measure the impact of design rules recommendations on the design process, we compare our approach with the use of traditional design rules documents. Results show that the use of the CACDA’s design rules recommendations lower the perceived difficulty of design rules retrieval from 75 to 43.5 on a scale of 100. On average, participants that used the demonstrator successfully applied 8.6 design rules on the 25 applicable design rules of our set. Participants that used unstructured documentation correctly applied 4.3 design rules. The global cognitive weight of the design activity as well as the design rules retrieval performances appear to be unchanged. These results demonstrate the usability of the demonstrator and show a positive impact on the design process and on the quality of CAD models. Future works will focus on the overcome of the main limitations of our current experiments, with a panel of professional designers, a lager design rules set and the implementation of several lacking features of the CACDA into the demonstrator.