Industry 4.0 features intelligent manufacturing. Among them, the vision-based defect inspection algorithm is remarkable for quality control in parts manufacturing. With the help of AI and machine learning, auto-adaptive instead of manual operation is achievable in this field, and much progress has been made in recent years. In this study, considering the demand of inspection features in industrialization, we made further improvement in smart defect inspection. An efficient algorithm using Field Programmable Gate Array (FPGA)-accelerated You Only Look Once (YOLO) v3 based on an attention mechanism is proposed. First, because of the relatively fixed camera angle and defect features, an attention mechanism based on the concept of directing the focus of defect inspection is proposed. The attention mechanism consists of three improvements: (a) image preprocessing, which is to tailor images for selectively concentrating on the defect relevant things. Image preprocessing mainly includes cutting, zooming and splicing, named CZS operations. (b) Tailoring the YOLOv3 backbone network, which is to ignore invalid inspection regions in deep neural networks and optimize the network structure. (c) Data augmentation. First, two improvements can be made to efficiently reduce deep learning operations and accelerate the inspection speed, but the preprocessed images are similar and the lack of diversity will reduce network accuracy. So, (c) is added to mitigate the lack of considerable amounts of training data. Second, the algorithm is deployed on a PYNQ-Z2 FPGA board to meet the industrialization production requirements for accuracy, efficiency and extensibility. FPGA can provide a low-latency, low-cost, high-power-efficiency and flexible architecture that enables deep learning acceleration for industrial scenarios. A Xilinx Deep Neural Network Development Kit (DNNDK) converted the improved YOLOv3 to Programmable Logic (PL), which can be deployed on FPGA. The conversion process mainly consists of pruning, quantization and compilation. Experimental results showed that the algorithm had high efficiency, inspection accuracy reached 99.2%, processing speed reached 1.54 Frames per Second (FPS), and power consumption was only 10 W.

This paper takes the industrial innovation ecosystem as the research object and the realization of green development as the goal, discussing the temporal and spatial evolution of coupling coordination degree of industrial innovation ecosystem from the perspective of system view. Based on the data of 30 provinces in China from 2010 to 2021, the spatial and temporal pattern distribution and spatial connection evolution of synergy among the three subsystems of industrial innovation ecosystem are studied by using coupling coordination degree model, trend surface model and gravity model. It is found that during the study period, the coupling relationship between the three subsystems is in a dynamic upward state. The regions with high values and rapid growth are distributed in the key areas of the national economic development strategy. At the same time, they have the characteristics of “positive U-shaped first and then inverted U-shaped” in the east-west and north-south directions; the spatial connection strength of coupling coordination degree shows that the spatial connection strength of the eastern region is significantly greater than that of the western region. With the improvement of spatial connection strength, a radial coupling network centered on Beijing Tianjin Tangshan region and the Yangtze River Delta is formed, which is of core significance to promote the coordinated development of industrial innovation ecosystem.

A key issue that concerns governments is how to formulate optimal technology subsidies and green tax standards to promote the intelligent transformation of manufacturing enterprises. In this work, the Pollutant Emission Indicator Trading Mechanism (PEITM) is proposed, and green taxes are divided into Tax of Pollutant Emissions (TPE) and the Tax of Excess Pollutant Emissions (TEPE). On this basis, we study the impact of green taxes and technology subsidies on the intelligent transformation of enterprises in different manufacturing environments from the government’s perspective and provide the optimal government subsidy scheme under different green tax policies. Although it seems counter-intuitive, enterprises’ usual responses to increases in TPE and TEPE are non-monotonic. Moreover, we find a threshold effect for the government’s green taxation. Blindly increasing or reducing taxes may not promote intelligent transformation, but instead force enterprises towards negative choices. Lastly, an effective measure for the government to promote the intelligent transformation of manufacturing enterprises is proposed: by properly adjusting TPE and TEPE, governments can produce more cost-effective intelligent products than ordinary ones. Consequently, intelligent products will sell better than ordinary products, and manufacturing enterprises will be able to consciously carry out intelligent transformation to remain viable.

Purpose: The purpose of this paper is to provide scientific guidance for the integration of industrialization and information (TIOII). In recent years, TIOII has promoted the development of intelligent manufacturing in China. However, many enterprises blindly invest in TIOII, which affects their normal production and operation. Design/methodology/approach: This study establishes an efficiency evaluation model for TIOII. In this paper, entropy analytic hierarchy process (AHP) constraint cone and cross-efficiency are added based on traditional data envelopment analysis (DEA) model, and entropy AHP–cross-efficiency DEA model is proposed. Then, statistical analysis is carried out on the integration efficiency of enterprises in Guangzhou using cross-sectional data, and the traditional DEA model and entropy AHP–cross-efficiency DEA model are used to analyze the integration efficiency of enterprises. Findings: The data show that the efficiency of enterprise integration is at a medium level in Guangzhou. The efficiency of enterprise integration has no significant relationship with enterprise size and production type but has a low negative correlation with the development level of enterprise integration. In addition, the improved DEA model can better reflect the real integration efficiency of enterprises and obtain complete ranking results. Originality/value: By adding the entropy AHP constraint cone and cross-efficiency, the traditional DEA model is improved. The improved DEA model can better reflect the real efficiency of TIOII and obtain complete ranking results.

Recently, car-parts suppliers all over the world are facing their margins under growing pressure due to intensified competition and the cost of innovation. In order to address these issues, most companies engage in redesigning production operations. This paper proposes an innovative approach consist of TRIZ and simulation technology, through which the unbalanced issues occurring in production lines could be found out, and suggestions for improvement can be put forward. The first step of this methodology framework is to define the technical problems using the FAA model. Secondly, a contradiction matrix is used for selecting tools resolving the problem potentially. Thirdly, the inventive principle are applied to generate the solutions against actual conditions. Fourthly, during the evaluation phase, the simulation software Witness and it’s optimization module are used to get an accurate flexible production schedule. In this study, an car-part supplier named HT company was taken as a case study; the comparative data showed that the company’s direct economic benefits is increased, the customer satisfaction and the workshop management level are improved.