Research on the construction and functions of intelligent food safety supervision systems

doi:10.21203/rs.3.rs-2220019/v1

Download PDF

Research Article

Research on the construction and functions of intelligent food safety supervision systems

https://doi.org/10.21203/rs.3.rs-2220019/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

This study investigated the existing problems and the optimal solutions in the construction of intelligent food safety supervision systems globally. Through browsing manually and searching by keywords about systems using the Internet as a medium, a total of 35 intelligent systems were collected and analyzed from the aspects of the servers, system performance, user groups, client construction, main functions, and characterized functions. The results showed that the current situation was a lack of unified data standards, the food safety risk early-warning model, sufficient system functional design and food safety professionals' participation, and systematic rules and regulations for system operation. Therefore, improvement measures including developing unified data standards, strengthening the cooperation and sharing of data resources, improving the information-sharing mechanism, building the multi-step dynamic risk early-warning models, and optimizing and improving the subfunctions and characteristic functions should be taken in the future.

Food safety supervision

Intelligent supervision

Risk assessment

Risk warning

Traceability system

Food safety supervision (FSS) is essential for ensuring food safety and avoiding food safety issues. FSS policies including the routine inspection policy and the risk-based grading policy are developed to protect consumers’ food safety by the government. Although many efforts have been made in FSS, food safety incidents happened frequently. For example, Salmonella outbreaks in the U.S. with 333 confirmed cases, including 91 hospitalizations in 2018 (Trinetta, et al., 2020). Listeria monocytogenes infected 1060 patients of whom 216 died in South Africa in 2019 (Boatemaa, et al., 2019). Recently, it was reported that clenbuterol was used illegally to raise sheep to seek a higher meat rate in China (Luo, et al., 2021). Specifically, some new food safety problems have arisen because the global COVID-19 epidemic intensifies. The COVID-19 virus was detected in the wholesale market salmon and ice cream samples, respectively (Lv, Xu, & Wang, 2021). According to World Health Organization (WHO), it was estimated that 600 million people fell ill through the consumption of contaminated food and 420,000 died every year, resulting in the loss of 33 million disability-adjusted life-years (WHO, 2015). Based on this situation, it is essential to apply new technologies and tools to improve the efficiency of FSS to maintain the health of consumers.

With the development of mobile Internet and information technology, some intelligent FSS systems were built and applied to daily FSS works (Chen, Ding, Yu, Li, & Dong, 2021). In the intelligent FSS systems, the big data analysis combined with the Internet platforms were used as an integrated solution that incorporated a food safety traceability system and food safety risk management system through Internet technology, Internet of Things technology, automatic identification technology, and automatic control technology. The advantages including the higher efficiency of FSS and the full-cycle supervision of the entire food safety industry chain had been proved in these intelligent FSS systems. Several studies have shown the importance of a well-functioning intelligent FSS system to prevent the risk of food safety hazards (Lunden, 2013; Osimani & Clementi, 2016; Pei, et al., 2011). Some countries such as the U.S. (Buchholz, Run, Kool, Fielding, & Mascola, 2002), the U.K, Finland (Kettunen, Pesonen, Lunden, & Nevas, 2018; Laikko-Roto, Makela, Lunden, Heikkila, & Nevas, 2015) had constructed intelligent FSS systems. Recently, the European Food Safety Authority constructed the Rapid Alert System for Food and Feed system, which had achieved significant results in controlling food safety issues (Martini, Del Bo, & Riso, 2020). It was worth mentioning that China had established some comprehensive and systemic intelligent FSS systems (Tang, et al., 2015), which had four typical models including the “regional synergy intelligent supervision model” (Han & Yan, 2019), the “problem-oriented intelligent supervision model” (Liu, Mutukumira, & Chen, 2019), the “multi-subject collaborative intelligent supervision model” (Donaghy, Danyluk, Ross, Krishna, & Farber, 2021) and the “full-cycle intelligent supervision model” (Lv, et al., 2021). Thus, based on the technologies of intelligence, network, and digital, intelligent FSS systems were widely used in daily supervision work and had proven their important value in reducing the occurrence of events damaging consumers’ health in China (Yu, et al., 2015). Nevertheless, the practices indicated that the intelligent FSS system was still in its beginning. On the one hand, the current intelligent FSS systems were designed and constructed by own understanding of IT people within the government, which resulted in a situation where different systems weren't synchronized with the business process and were difficult to integrate supervision information. On the other hand, little research had been made on the components and functions of the intelligent FSS system. Besides, the standards on the construction and functions of intelligent FSS systems haven’t been developed up to now.

Therefore, there was a lack of research on the frameworks and functions of intelligent FSS systems, which impeded their construction and improvement. To find the existing problems and optimal solutions, it is necessary to systemically analyze the functions and conditions of intelligent FSS systems globally. Detailed investigations were conducted in the following section. Firstly, the basic information about the intelligent FSS systems was collected from the website of the agricultural department and the FSS department. Secondly, all systems’ information was analyzed from the aspects of the servers, system performance, user groups, client construction, main functions, characterized functions, and so on. Finally, suggestions were given from the users, hosts, and governments. And the optimal solutions were proposed to FSS departments, which will be of great significance for improving the future development and management of FSS.

2.1. Data Collection

Data collection of intelligent FSS systems was carried out from the website and covered 34 regions around the world. The last collection time was by the end of August 2022. The procedure of collection had been conducted in the following steps. Firstly, the official websites of FSS departments were found by searching the keywords of “regions name” and “market regulation department” or “agricultural administration” using Microsoft Edge. Secondly, according to the internet site (URL) of the searched results, the intelligent FSS systems were browsed manually and searched using keywords such as “food safety supervision”, “food safety supervision system”, or “food safety supervision platform” on the webpage. Thirdly, the operation manuals and introductions of the FSS systems were acquired on the home page.

All collected information about the FSS systems was obtained from official sources. The information on the system equipment was collected from the operation manual and introduction. The function information was obtained by browsing and using them after logging in to the FSS systems. If the FSS system couldn’t be logged in, all information was obtained from the system technical supporter by email or telephone. Besides, the information was made up of news or literature.

2.2. Data treatments

A total of 75 intelligent FSS systems were found including 68 systems located in China and 7 systems in other countries. Due to system login authority and privacy issues, the information was collected from 35 intelligent FSS systems. The names of the systems that collected the data and affiliations were shown in Table 1.

Table 1

Name of food safety supervision systems and affiliations
No.	System name	Affiliation	Country
1	Anhui cold chain food traceability platform	Market Regulatory Administration of Anhui Province	China
2	Beijing cold chain food traceability platform	Market Regulatory Administration of Beijing	China
3	Fujian food safety information traceability management platform	Market Regulatory Administration of Fujian Province	China
4	Gansu food safety information traceability platform	Market Regulatory Administration of Gansu Province	China
5	Guangdong food traceability platform	Market Regulatory Administration of Guangdong Province	China
6	Hainan trusted traceability platform for cold chain food	Market Regulatory Administration of Hainan Province	China
7	Hebei cold chain food traceability management system	Market Regulatory Administration of Hebei Province	China
8	Henan agricultural product quality and safety intelligent supervision and full traceability platform	Department of Agriculture and Rural Affairs of Henan Province	China
9	Henan province cold chain food traceability system	Market Regulatory Administration of Henan Province	China
10	Hubei imported cold chain food information traceability platform	Market Regulatory Administration of Hubei Province	China
11	Hunan imported cold chain food traceability system	Market Regulatory Administration of Hunan Province	China
12	Imported food traceability and risk early warning platform	Shenzhen Customs	China
13	Inner Mongolia autonomous region imported cold chain food traceability platform	Market Regulatory Administration of Inner Mongolia Autonomous Region	China
14	Jiangsu province imported cold chain food traceability system	Market Regulatory Administration of Jiangsu Province	China
15	Jiangxi food safety traceability platform	Market Regulatory Administration of Jiangxi Province	China
16	Jilin imported cold chain food traceability management system	Market Regulatory Administration of Jilin Province	China
17	Liaoning agricultural product quality safety supervision and service platform	Department of Agriculture and Rural Affairs of Liaoning Province	China
18	Hunan province application system of national agricultural products quality and safety traceability platform	Department of Agriculture and Rural Affairs of Hunan Province	China
19	National animal identification and tracing	Ministry for Primary Industries	New Zealand
20	Ningxia global food safety traceability platform	Market Regulatory Administration of Ningxia Hui Autonomous Region	China
21	Shaanxi agricultural products quality and safety supervision platform	Department of Agriculture and Rural Affairs of Shaanxi Province	China
22	Shandong cold chain food epidemic prevention and control management system	Market Regulatory Administration of Shandong Province	China
23	Shanghai food safety information traceability platform	Market Regulatory Administration of Shanghai	China
24	Shanxi agricultural products quality and safety supervision traceability information platform	Department of Agriculture and Rural Affairs of Shanxi Province	China
25	Sichuan cold chain food traceability platform	Market Regulatory Administration of Sichuan Province	China
26	The Canadian livestock tracking system	Canadian Food Inspection Agency	Canada
27	The nation animal identification system	United States Department of Agriculture	USA
28	The national livestock identification system	Integrity Systems Company	Australia
29	The rapid alert system for food and feed	Directorate-General for Health and Food Safety	European Union
30	Tianjin cold chain food traceability platform	Market Regulatory Administration of Tianjin	China
31	Traceability scheme for cow/beef	Ministry of Agriculture, Forestry, and Fisheries	Japan
32	Traceability system for rice/rice products	Ministry of Agriculture, Forestry, and Fisheries	Japan
33	Xinjiang agricultural product quality and safety traceability management platform	Department of Agriculture and Rural Affairs of Xinjiang Province	China
34	Yunnan cold chain food safety information traceability platform	Market Regulatory Administration of Yunnan Province	China
35	Zhejiang cold chain food traceability system	Market Regulatory Administration of Zhejiang Province	China

To ensure data accuracy, all FSS systems collected in the present study need to be put into service. The information can’t be collected from one system which is the agricultural products quality and safety traceability platform of Guangdong province in China (http://qsst.gdagri.gov.cn) because it hadn’t been put into service. Moreover, the cold chain food traceability system of Zhejiang province and the food safety traceability closed-loop management system of Zhejiang province in China had no significant differences in function and construction, which were both constructed by the same technology company. One of two systems was selected as a representative for information statistics.

2.3. Data analysis

The information of the intelligent FSS system was summarized and converted to quantitative data in an Excel table (Microsoft Excel 2019, Redmond, USA). Based on these data, digital forms and figures were made using Microsoft Excel 2019 (Microsoft, Redmond, WA, USA). Two distribution maps were created in Python (version 3.7.2) using Anaconda (version 4.10.3, https://www.anaconda.com/, Anaconda Inc., Austin, TX, USA) and Jupyter notebook (version 6.4.5, https://jupyter.org).

3.1. Distribution Analysis

The distribution of the intelligent FSS systems in China and the world was shown in Fig. 1. According to Fig. 1A and Fig. 1B, the intelligent FSS systems were built in the most province in China, while they were set up in a few countries in the world. It might be associated with the strategies adopted by the different countries. Huge financial and human investments were needed for the new intelligent FSS systems. In China, there was a more powerful government tried to have the best regulatory efficiency by taking the new intelligent FSS measures. However, other countries were satisfied with the current FSS strategies, which were thought to be enough to guarantee food safety.

3.2. Analysis of server and system performance

The servers of the intelligent FSS systems were associated with the development of the Network. Cloud computing servers were used in most of the FSS systems. For example, the Alibaba Cloud servers were selected by the Hunan imported cold chain food traceability system, Zhejiang cold chain food traceability system, and the Hubei imported cold chain food information traceability platform. Meanwhile, a physical host server was applied to FSS systems due to the advantages of stable performance and higher security. According to the survey, only one physical host’s server was used by the Imported food traceability and risk warning system of Shenzhen customs. Its physical host was built by 2 Xeon Silver 414 10-core CPUs (2.2GHz) and 128 Gigabytes of memory. Up to now, cloud computing servers were commonly used in FSS systems because they had the advantages of rapid deployment, elastic storage, and low cost. The efforts and promotions made by cloud service providers also was another reason. With the fast development of cloud servers, cloud computing servers were widely applied to the FSS system. Moreover, there are many options for cloud service providers including Bluehost, HostGator Cloud, GreenGeeks, Alibaba Cloud, HUAWEI CLOUD, Tencent Cloud, and BAIDU AI CLOUD. Although the different requirements of the different FSS systems were probably the reasons for the inconsistency of server selections, a cloud computing server was the optimal choice for the FSS system.

Both cloud computing servers and physical host servers can provide enough processing capability and performance for the FSS systems. The typical and basic requirements for processing capability performance are as follows. The enterprise end of the FSS systems can support 1,000 users, 200 simultaneous online users, and 20 concurrent users, respectively. The supervisory end of the FSS systems can support 300 users, 30 simultaneous online users, and 5 concurrent users, respectively. For the FSS systems response time, the response time of transaction processing events ≤ 3 seconds, the response time of information query events ≤ 5 seconds, and the response time of statistical report events ≤ 7 seconds are necessary. One Gbps was allocated in most of the FSS systems’ network bandwidth. Some FSS systems also were equipped with more than 1 Gbps. For example, 6 ethernet ports of Gbps and 2 optical ports of 10 Gbps were used by the Imported food traceability and risk warning system of Shenzhen customs.

3.3. User groups analysis

The composition of user groups for each FSS system was shown in Table 2. The users of the FSS system were divided into three groups including government supervisors, food enterprises, and consumers. The government supervisors contained agricultural departments, FSS departments, and market management departments, which were responsible for FSS works. Food enterprises had more complicated sub-groups including food producers, food processors, food distributors, supermarkets, restaurants, and so on. Three user groups were provided services by 32 FSS systems except for the Hunan imported cold chain food traceability system, Zhejiang cold chain food traceability system, and the Hubei imported cold chain food information traceability platform. There was no consumer group in these three FSS systems.

Table 2

The composition of user groups
User groups	Number
Government supervisors	35
Food enterprises	35
Consumers	32

User groups were the input or output ports of information in the FSS system. Normally, the FSS system started from a group of food enterprises, which input the digital information of food traces and risks into the system. All groups queried food traceability information through the FSS system. The supervisor’s group played an FSS role in the food circulation from the aspect of the whole chain traceability. Therefore, the three user groups had different functions and responsibilities in the FSS systems.

3.4. Client construction analysis

Both the web browser client and mobile client were used in the FSS systems. The web browser client was worked by web browsers including Firefox, Google Chrome, Safari, and Internet Explorer. The web browser client is rendered within the browser window, which can transmit a service request sent by a user to a server. Web browser clients can run a software application or a web on PC or mobile devices. The mobile client is divided into the application (APP) and the WeChat Mini Program in the FSS systems. In practice, mobile clients are popular with the development of mobile communications.

Table 3

Client construction statistics
Client	Number
Web	34
WeChat Mini Program	11
APP	15

The client construction of the FSS systems was shown in Table 3. Except for the Inner Mongolia autonomous region imported cold chain food traceability platform, the web browser clients were constructed in all FSS systems. Meanwhile, in terms of the mobile client, the APPs were developed by 15 FSS systems, whereas the WeChat Mini Programs were built by 11 FSS systems. Further analysis of the clients was shown in Fig. 2. The web browser client, APP, and WeChat Mini Program were constructed in 3 FSS systems. The web browser client and APP were constructed in 12 FSS systems. The web browser client and WeChat Mini Program were constructed in 7 FSS systems.

The web browser client was extensively constructed in the FSS systems due to three advantages of mature technology (Dong, Liu, Han, & Yang, 2016), without additional software or plug-ins (Boyington, Wildemuth, Dougherty, & Hall, 2004), and cross-platform compatibility (Bessghaier, Ouni, & Mkaouer, 2021). With the popularity of mobile devices, the mobile client was widely used in the FSS system. Based on the advantages of personalized customization, flexibility, and a large user base, APPs were common in the construction of the FSS system. Meanwhile, the WeChat Mini Program was not only favored by the users of the FSS system because of direct access from the WeChat search list without downloading and installing operations but also was accepted by the developers due to the light development and low threshold.

3.4. The analysis of main functions

The functional architecture of the FSS systems was shown in Fig. 3. According to the business of the intelligent FSS systems, the main functions were classified into food traceability, food safety risk warning, and system management. The main functions were divided into many subfunctions to be put into service efficiently and accurately. For example, the subfunctions in food traceability included logistics distribution information management, sales information management, regulatory management, and consumer inquiries. The functions were presented graphically and effectively to users through information visualization techniques. For example, visualization technologies were used to analyze food safety risk assessment results and food safety risk traceability processes. Migration maps and force-directed graphs were also applied to show microscopic flow. The users easily tracked and monitored the extent and impact of the spreading of the dangerous food using a human’s rapid recognition ability in visual mode (Hao, Mao, Zhang, Zuo, & Zhao, 2020).

In the FSS systems, the development of subfunctions was similar in the different clients. The development of three main functions was analyzed and shown in Fig. 4A. The food traceability function and the system management function were developed in all FSS systems. However, the risk warning function was only constructed in 16 systems, which accounted for 45.7% of the collected FSS systems.

3.4.1. Food traceability function

The food traceability function was composed of six subfunctions including traceable resource unit, food information collecting and management, sales information management, supervision management, consumer query, and problematic product management. These subfunctions worked synergistically to guarantee the normal operation of the main function.

Among them, the traceable resource unit was the basic element of the food traceability function. Traceable resource units were given unique identities using coding technologies such as barcodes or radio frequency identification (RFID), which were essential to individually track a traceable resource unit through the supply chain. In the subfunction of food information collecting and management, the food information in the supply chain could be queried, added, modified, or deleted by government supervisors and food enterprises. In the subfunction of supervision management, government supervisors can check and view all information in the supply chain. In the subfunction of the consumer query, consumers not only can query the information including origin, production, processing, and transportation using the unique identification code on the packaging bag but also can complain about products by providing complaint messages. In the subfunction of problematic product management, government supervisors can implement product inspection. A problematic product record will be generated if the product is unqualified in the inspection.

The food traceability function was analyzed and shown in Fig. 4B. Three subfunctions including traceable resource unit, food information collecting and management, and supervision management were developed in all FSS systems. It was suggested that these subfunctions were indispensable in a food traceability function. The sales information management, the consumer query, and the problematic product management subfunction were developed by 75%, 18%, and 24% of the FSS systems, respectively. The low proportion of consumer queries and problematic product management subfunctions might be associated with the aims and users of FSS systems. Few consumers were involved in the FSS system because the consumer query was the only way for consumers, which indicated that the rights and interests of consumers were neglected. Thus, the construction of the food traceability function was focused on the traceability of food information and movement.

In the subfunction of the traceable resource unit, different dictionaries were used to encode the food name and the composition of the traceable resource unit. According to Fig. 5A, GS1 standards, enterprise codes, Harmonized System (HS) codes, and numbering system identifier (NSI) were used to encode the food name by 53%, 25%, 18%, and 4% in the FSS systems, respectively. GS1 standards were widely used in the field of global circulation because GS1 standards were international. While the exorbitant prices of GS1 standards might lead a certain proportion of the FSS systems to use enterprise codes. Moreover, the low proportion of HS codes might be associated with the types of FSS systems. The HS codes were normally used in the FSS systems which correlated with the import and export trade. Furthermore, the lowest proportion of NSI might be related to its native standard in China.

The composition of the traceable resource unit was shown in Fig. 5B. The customized traceable resource unit was used in 11% of FSS systems, while the traceable resource unit of "GS1 barcodes + batch production" mode was used in 82.1% of FSS systems. Additionally, both ways were used in 7% of FSS systems. The composition of the traceable resource unit was associated with the types of the dictionary, and the composition of the traceable resource unit in most FSS systems was based on the dictionary. On the one hand, the dictionary of GS1 standards was wildly used in the FSS systems. On the other hand, imported goods that used the HS code were transformed coding into GS1 standards, when they circulated in the market. The part of FSS systems that used enterprise codes also choose the GS1 standards as commodity bar codes. So the combination of GS1 barcodes and batch production was wildly used in the composition of the traceable resource unit. The customized traceable resource unit was used in the rest FSS systems because of is convenient that the codes don‘t need to be transformed. At the present, the Thing technology combined with coding technologies such as article coding and RFID had been used in the food traceability function to manage information flow in the food supply chain, which was the assurance of food safety in the FSS systems (Zheng, Zhang, Zhang, & Hu, 2021).

3.4.2. Risk warning function

The risk warning function plays a role in the early-warning and prevention of food safety, which can provide a mechanism to reduce the risk of food safety incidents through the monitoring of food safety problems, tracking, quantitative analysis, and forecast information (G. Liu, Li, Yang, & Guo, 2018). The risk warning function was composed of four subfunctions including risk information acquisition, risk warning model, early-warning analysis, and early-warning alert. The subfunction of risk information acquisition included the selection of the key points of food safety regulation, the discovery of the hot spots of public opinion, and the sharing of big data resources. The risk information was collected from the media, government organizations, and non-governmental organizations, such as the WHO, the food safety supervision department, newspapers, and internet public opinion. In the subfunction of the risk warning model, the development trend of food safety risk was predicted based on the monitoring and analysis of the dynamic changes of risk information. In the subfunction of the early-warning analysis, the predicted risk information will be evaluated and classified into the risk early-warning levels, which are determined according to the risk value distribution. When the predicted risk values are within the high-risk levels, the risk may be in the level of danger, which means a high possibility of occurrence. In the subfunction of the early-warning alert, once the risk early-warning level is reached, the early-warning information will be generated and automatically pushed to the government supervisors for further screening and judgment.

The risk warning subfunction was analyzed and shown in Fig. 4C. The early-warning alert was developed by 100%, which was the core of the risk warning function. The subfunction of risk information acquisition and early-warning analysis was developed in 38% of the risk warning function. The technical limitations are the main reason for the partial development of these subfunctions. The risk information acquisition used a web crawler which is an automatic extraction of web pages to collect the risk information from the web pages. The data processing model was required by the subfunction of early-warning analysis. In 19% of the risk warning function, the subfunction of the risk warning model was developed. The low proportion of the risk warning model subfunction might be associated with the lack of risk warning models. The development of the risk warning models was difficult for most governments or enterprises. Otherwise, the risk warning objects in some systems were thought to be objective factors such as production date, the sanitation status of producers and distributors, and other factors, which don’t need the risk warning model to analyze.

3.4.3. System management function

The system management function guarantees the stable operation of the system functions based on comprehensive regulation of the information, which is composed of four subfunctions including basic information management, roles management, rights management, and audit log management. In the subfunction of basic information management, the information of the enterprises and government supervisors was refined with inventory management, warehouse management, certificate management, and so on. The subfunction of roles management establishes various operating users for the system and manages users. In the subfunction of rights management, users are authorized to access different resources based on their role in the system. In the subfunction of audit log management, the logs yielded by the hardware and software such as routers, switches, servers, and databases were collected and stored, which will help the system administrator to check the operation of the system and to detect anomalous behavior in the systems.

Subfunctions in the system management were analyzed and shown in Fig. 4D. Three subfunctions including basic information management, roles management, and rights management were developed in all FSS systems. The basic information was filled in at registration, which is the necessary condition to log into the system for the enterprises and government supervisors. The subfunction of roles management defines the role and permissions of the users in the systems, which is associated with the rights management subfunction. A principle to construct the systems is to provide exclusive operating systems for different groups of users. Thus, the subfunction of roles management and rights management was developed in all systems. In addition, the low proportion of the audit log management subfunction might be associated with access rights. The audit log management subfunction was developed only for the system administrators, it usually can’t be found on the desktop page for the systems.

3.5. Characterized functions analysis

Except for the mentioned main functions, some characterized functions also had been constructed in the FSS systems. The reason for constructing these characterized functions was summarized in the following three points. Firstly, professional management functions were required for the supervision of the special foods. For example, fresh food and frozen food require the cold-chain management function to the maintenance of food quality and safety in the supply chain. Secondly, to meet the new regulatory requirements, some were proposed by the government’s regulatory bodies based on the social situation. For example, in the context of COVID-19, the government stipulated that workers for food production and transportation had to provide negative nucleic acid results within 24 hours or seven days, which resulted in the construction of the report management function of nucleic acid test in the FSS systems. Thirdly, the characterized function also included some new ways to improve the FSS efficiency. For example, in the Anhui Province Cold Chain Food Traceability Platform, the characterized function of enterprise self-examination was constructed to prevent unqualified products from the market. In the Hunan Province Imported Cold Chain Food Traceability System, the characterized function of credibility management was constructed to credit ratings for enterprises according to their scores. Government supervisors can strengthen regulations depending on the credit rating of enterprises. In the Shenzhen Imported Food Traceability and Risk Early Warning Platform, the characterized function of the food safety profile was constructed to display the information in the form of graphs and tables, which help the government supervisors and food enterprises have a clear understanding of the risk information.

An intelligent FSS system integrates advanced management technologies with food safety supervision and uses modern information technology to automatically collect information about food safety (Song, Yu, & Lv, 2018). However, the current construction of the FSS systems cannot fully fulfill the requirements of the FSS development (Derpmann, 2018). The difficulties in the actual operation of intelligent food safety supervision mainly focused on the following problems.

4.1. Non-unified data standards

Many FSS systems were constructed by different institutions using different technologies, which caused the non-unified data standards. For example, different dictionaries were used to encode the food name in the FSS systems of the different institutions. The Customs used the HS code while the regulatory authorities used the GS1 standard code, which may cause mismatched information between the systems in the process of food circulation. Besides, the traceable resource unit is not uniform in the food traceability function which increases the difficulty of traceability. For example, the systems used different types of traceable resource units. Most were QR codes while some were barcodes. The QR codes can be directly recognized by mobile phones, but the recognition of barcodes requires special equipment, which made the traceability process difficult.

4.2. Missing the multi-step dynamic risk early-warning model

Common early-warning models of food safety risk were based on grey relational analysis (Hashemi, Karimi, & Tavana, 2015) and artificial neural networks analysis (Horiuchi, et al., 2004), which were applied to food safety prediction problems. Recently, neural network analysis was applied to food safety early-warning models to improve the prediction precision (Geng, et al., 2021). However, typical shallow neural network models are single-step predictions, which causes the models cannot predict unknown data or future occurrences. The food safety risk prediction requires the early-warning models to grasp the long-term change trend of food safety risks and to realize an accurate multi-step prediction of the risks. Meanwhile, to improve regulatory efficiency, risk levels need to be assessed and the different risk levels have to be prioritized differently by the early-warning models. Moreover, the early-warning models should identify the causes of each risk level and evaluate the effect of each index on the classification so that risk management can achieve a retrospective analysis of food safety risks and develop targeted strategies.

4.3. Unreasonable and unprofessional function design

The establishment of the FSS system is a multidisciplinary and highly professional work (Gallo, Accorsi, Goh, Hsiao, & Manzini, 2021), which requires the participation of professionals who know about food safety, food toxicology, food safety risk assessment, as well as professionals with software programming, platform development, network technology, data information visualization, and big data. However, most FSS system developers couldn’t closely integrate digital technology into the FSS business because they had no professional food safety knowledge. This led to the illogical design of some FSS systems. At the beginning of the design, the developers usually pursued comprehensive functions but ignored the operability and convenience of the systems. For example, due to inaccurate preliminary research and incomplete supporting software, some daily office modules of the big data processing system were not put into use after the system was installed. And some subfunctions of food safety supervision became idle due to problems such as cumbersome operations and too many approving personnel (Chen, et al., 2021).

4.4. No systematic rules and regulations for system operation

Nothing can be accomplished without norms or standards. If a reasonable set of rules and regulations can be formulated scientifically and put in place, it will be more effective. However, it is difficult to fundamentally solve the dilemma faced by FSS without relevant laws supporting it due to the following points. Firstly, the data of the FSS system needs to be accurate and safe because most of the systems were used by the government, which requires regulators to have basic professionalism. Once the regulators input the wrong information or cause the database to be leaked, they should bear legal responsibility. However, no corresponding laws have been established yet (He, Yan, & Chen, 2018).

Secondly, the authenticity of the information in the corporate entry system cannot be guaranteed due to the lack of relevant regulatory laws (Marques & Reis, 2020). Enterprises must be completely transparent about regulatory information, which needs support from the laws. Moreover, there aren’t standards for the preservation of supervisory information in the FSS systems such as the preservation time of surveillance videos and the disclosure time of corporate ledger information. Additionally, the FSS system requires enterprises to clarify the principal responsibility that the enterprise shall bear responsibility for the authenticity of regulatory information and the food quality and risk in the whole process of food production.

Thirdly, the system needs to protect the privacy of corporate information and consumer personal information. The leakage of corporate information may cause economic losses to the corporate in the competition (Kau & Kogeda, 2019). The leakage of consumers' personal information may cause them to receive sales calls, harassing information, spam, etc. (Chung, Chen, Tsai, & Chuang, 2021). Therefore, it is necessary to have regulations to regulate the security of the systems and the behavior of regulators.

5.1. Promote the sharing of information between departments

Ignoring the communication and coordination between departments will inevitably lead to the problem of information islands. The key to solving the information islands lies in strengthening the management of government and improving the information-sharing mechanism between departments (K. L. Wang & Cao, 2010). The basis for sharing is to establish unified data standards and to establish a systematic enterprise information database by integrating information about the systems between government departments (Zheng, et al., 2021).

5.2. Build the multi-step dynamic risk early-warning model

The multi-step dynamic risk early-warning model can enable regulators to measure food safety risks promptly and address future problems. Using the time series forecasting model combined with the deep learning method based on the historical data, FSS systems not only can predict future operation patterns but also can adopt a voting-ensemble method to increase the accuracy of the prediction model (Wang, et al., 2021). The model can efficiently capture the dynamic operation pattern of the food safety incidents time series and improve early-warning accuracy for food safety. Meanwhile, the models can strengthen the risk assessment capacity of food production information such as reporting the risks that do not meet safety requirements or inaccurate labeling (Donaghy, et al., 2021).

5.3. Optimize the function design

In the function design of the FSS systems, the developers should focus on system integration and function optimization by unifying data standards and upgrading function modules. It also was important to find out the system bugs, idle modules, and impractical modules which are encountered by all user groups during the use of these FSS systems. Moreover, the developers should not only focus on function optimization but also consider the comprehensiveness of the system, that is the whole production process and the whole industry chain involved in FSS. The business functions of the systems should be designed according to the process of food from entering the market to the consumer. In addition, the overlapping information should be detected and deleted automatically to simplify systems.

The intelligent FSS system had played an important role in guaranteeing food safety in the daily FSS works. Our results suggested that China had led the development of an intelligent FSS system associated with government strategies. The cloud computing server was the optimal choice for the FSS system. Three user groups including government supervisors, food enterprises and consumers, and both clients including the web browser and mobile were used in all FSS systems. The food traceability, food safety risk assessment and warning, system management functions, and other characterized functions were designed in the current FSS systems. Nevertheless, the FSS system was still faced with the challenges of non-unified data standards, missing food safety risk early-warning models, insufficient functional design of systems and food safety professionals' participation, and no systematic rules and regulations for system operation. To achieve the smart and intelligent FSS, the recommendations including developing unified food safety data standards, strengthening the cooperation and sharing of data resources, and improving the information-sharing mechanism had been put forward. It also is important to build the multi-step dynamic risk early-warning models and to optimize and improve subfunctions and characteristic functions.

Acknowledgments

This research was funded by the Basic Research Program of Shenzhen Municipal Government (JCYJ20200109114242138), the National Natural Science Foundation of China (22193064, 22078198), the Shenzhen Science and Technology Program (ZDSYS20210623100800001), the Natural Science Foundation of Guangdong Province (2021A1515010687) and the Science and Technology Commissioner for Rural Revitalization in Guangdong (KTP20210345).

Consent to participate

Informed consent was obtained from all individual participants included in the study.

Consent to publish

The participant has consented to the submission of the case report to the journal.

Funding

Competing interests

The authors have no competing interests to declare that are relevant to the content of this article.

Data Availability Statements

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors’ contribution statements

Conceptualization: Guanzhao Wang, Shijia Li, Yongning Wu; Methodology: Guanzhao Wang, Shijia Liu, Yajie Wang, Jing Cao, Pingping Ren; Formal analysis: Guanzhao Wang; Investigation: Guanzhao Wang, Yajie Wang, Jing Cao, Pingping Ren; Data curation: Guanzhao Wang; Resources: Shijia Liu, Qinghua He; Writing – original draft: Guanzhao Wang; Writing – review & editing: Yongning Wu, Qinghua He; Supervision: Yongning Wu, Qinghua He.

Competing Interests

The authors declare no conflict of interest.

Bessghaier, N., Ouni, A., & Mkaouer, M. W. (2021). A longitudinal exploratory study on code smells in server side web applications. Software Quality Journal, 29(4), 901–941. https://doi.org/10.1007/s11219-021-09567-w
Boatemaa, S., Barney, M., Drimie, S., Harper, J., Korsten, L., & Pereira, L. (2019). Awakening from the listeriosis crisis: Food safety challenges, practices, and governance in the food retail sector in South Africa. Food Control, 104, 333–342. https://doi.org/10.1016/j.foodcont.2019.05.009
Boyington, A. R., Wildemuth, B. M., Dougherty, M. C., & Hall, E. P. (2004). Development of a computer-based system for continence health promotion. Nursing Outlook, 52(5), 241–247. https://doi.org/10.1016/j.outlook.2004.04.014
Buchholz, U., Run, G., Kool, J. L., Fielding, J., & Mascola, L. (2002). A risk-based restaurant inspection system in Los Angeles County. Journal of Food Protection, 65(2), 367–372. https://doi.org/10.4315/0362-028x-65.2.367
Chen, T. B., Ding, K. F., Yu, Z. Y., Li, G. D., & Dong, Y. I. (2021). Smart supervision for food safety in food service establishments in China: Challenges and solutions. Journal of Food Protection, 84(6), 938–945. https://doi.org/10.4315/Jfp-20-370
Chung, K. C., Chen, C. H., Tsai, H. H., & Chuang, Y. H. (2021). Social media privacy management strategies: A SEM analysis of user privacy behaviors. Computer Communications, 174, 122–130. https://doi.org/10.1016/j.comcom.2021.04.012
Derpmann, S. (2018). Union's inspiration: Universal health care and the essential partiality of solidarity. Bioethics, 32(9), 569–576. https://doi.org/10.1111/bioe.12444
Donaghy, J. A., Danyluk, M. D., Ross, T., Krishna, B., & Farber, J. (2021). Big data impacting dynamic food safety risk management in the food chain. Frontiers in Microbiology, 12, 668196. https://doi.org/10.3389/fmicb.2021.668196
Dong, Y. S., Liu, Y. Q., Han, X. B., & Yang, J. (2016, September). Design of multi-terminal mobile learning platform. Fifth International Conference on Educational Innovation through Technology, Taiwan, China., 37–41. https://doi.org/10.1109/Eitt.2016.15
Gallo, A., Accorsi, R., Goh, A., Hsiao, H., & Manzini, R. (2021). A traceability-support system to control safety and sustainability indicators in food distribution. Food Control, 124, 107866. https://doi.org/10.1016/j.foodcont.2021.107866
Geng, Z. Q., Liu, F. F., Shang, D. R., Han, Y. M., Shang, Y., & Chu, C. (2021). Early warning and control of food safety risk using an improved AHC-RBF neural network integrating AHP-EW. Journal of Food Engineering, 292, 110239. https://doi.org/10.1016/j.jfoodeng.2020.110239
Han, G., & Yan, S. (2019). Does food safety risk perception affect the public's trust in their government? An empirical study on a national survey in China. International Journal of Environmental Research and Public Health, 16(11), 1874. https://doi.org/10.3390/ijerph16111874
Hao, Z. H., Mao, D. H., Zhang, B., Zuo, M., & Zhao, Z. H. (2020). A novel visual analysis method of food safety risk traceability based on blockchain. International Journal of Environmental Research and Public Health, 17(7), 2300. https://doi.org/10.3390/ijerph17072300
Hashemi, S. H., Karimi, A., & Tavana, M. (2015). An integrated green supplier selection approach with analytic network process and improved grey relational analysis. International Journal of Production Economics, 159, 178–191. https://doi.org/10.1016/j.ijpe.2014.09.027
He, H. W., Yan, A., & Chen, Z. H. (2018). Survey of smart contract technology and application based on blockchain. Journal of Computer Research and Development, 55, 2452–2466. https://doi.org/10.7544/issn1000-1239.2018.20170658
Horiuchi, J., Shimada, T., Funahashi, H., Tada, K., Kobayashi, M., & Kanno, T. (2004). Artificial neural network model with a culture database for prediction of acidification step in cheese production. Journal of Food Engineering, 63(4), 459–465. https://doi.org/10.1016/j.jfoodeng.2003.09.005
Kau, F. M., & Kogeda, O. P. (2019, October). Impact of subscription fraud in mobile telecommunication companies. 2019 Open Innovations Conference, Cape Town, South Africa., 42–47. https://doi.org/10.1109/OI.2019.8908261
Kettunen, K., Pesonen, S., Lunden, J., & Nevas, M. (2018). Consistency and risk-basis of using administrative enforcement measures in local food control. Food Control, 85, 199–211. https://doi.org/10.1016/j.foodcont.2017.09.023
Laikko-Roto, T., Makela, S., Lunden, J., Heikkila, J., & Nevas, M. (2015). Consistency in inspection processes of food control officials and efficacy of official controls in restaurants in Finland. Food Control, 57, 341–350. https://doi.org/10.1016/j.foodcont.2015.03.053
Liu, G., Li, G., Yang, R., & Guo, L. (2018, September). Improving food safety in supply chain based on big data. In 2018 3rd International Conference on Advances in Energy and Environment Research. 53, 03084. https://doi.org/10.1051/e3sconf/20185303084
Liu, Z., Mutukumira, A. N., & Chen, H. (2019). Food safety governance in China: From supervision to coregulation. Food Science & Nutrition, 7(1), 4127–4139. https://doi.org/10.1002/fsn3.1281
Lunden, J. (2013). Reasons for using enforcement measures in food premises in Finland. Food Control, 31(1), 84–89. https://doi.org/10.1016/j.foodcont.2012.09.046
Luo, Y. Y., Chen, L. J., Xu, F., Gao, X. X., Han, D., & Na, L. X. (2021). Investigation of knowledge, attitudes and practices about food safety and nutrition in China during the epidemic of corona virus disease 2019. Public Health Nutrition, 24(2), 267–274. https://doi.org/10.1017/S1368980020002797
Lv, C., Xu, J., & Wang, C. (2021). Intelligent strategies to improve food safety supervision model. International Journal of Education and Humanities, 1(1), 21–27. https://doi.org/10.54097/ijeh.v1i1.197
Marques, A., & Reis, L. (2020, June). QR-Code on tax-relevant documents in Portugal. 2020 15th Iberian Conference on Information Systems and Technologies, Barcelona, Spain., 1–7. https://doi.org/10.23919/CISTI49556.2020.9140819
Martini, D., Del Bo, C., & Riso, P. (2020). The European legislation and the role of ESFA in safety evaluation: Principles and perspectives. Biolaw Journal-Rivista Di Biodiritto, 2, 9–23. https://doi.org/10.15168/2284-4503-647
Osimani, A., & Clementi, F. (2016). The occurrence of Listeria monocytogenes in mass catering: An overview in the European Union. International Journal of Hospitality Management, 57, 9–17. https://doi.org/10.1016/j.ijhm.2016.05.005
Pei, X. F., Tandon, A., Alldrick, A., Giorgi, L., Huang, W., & Yang, R. J. (2011). The China melamine milk scandal and its implications for food safety regulation. Food Policy, 36(3), 412–420. https://doi.org/10.1016/j.foodpol.2011.03.008
Song, Y. H., Yu, H. Q., & Lv, W. (2018). Risk analysis of dairy safety incidents in China. Food Control, 92, 63–71. https://doi.org/10.1016/j.foodcont.2018.04.007
Tang, Q., Li, J. J., Sun, M., Lv, J., Gai, R. Y., Mei, L., & Xu, L. Z. (2015). Food traceability systems in China: The current status of and future perspectives on food supply chain databases, legal support, and technological research and support for food safety regulation. Bioscience Trends, 9(1), 7–15. https://doi.org/10.5582/bst.2015.01004
Trinetta, V., Magossi, G., Allard, M. W., Tallent, S. M., Brown, E. W., & Lomonaco, S. (2020). Characterization of Salmonella enterica isolates from selected US swine feed mills by whole-genome sequencing. Foodborne Pathogens and Disease, 17(2), 126–136. https://doi.org/10.1089/fpd.2019.2701
Wang, K. L., & Cao, D. R. (2010, September). A new path explores city's digit management. 2010 6th International Conference on Wireless Communications Networking and Mobile Computing, Chengdu, China., 1–4. https://doi.org/10.1109/WICOM.2010.5601214
Wang, M. J., Li, S. L., Chen, Z. J., Zhu, J. B., Hao, W. D., Jia, G., Chen, W., Zheng, Y. X., Qu, W. D., & Liu, Y. (2021). Safety assessment of nanoparticles in food: Current status and prospective. Nano Today, 39, 101169. https://doi.org/10.1016/j.nantod.2021.101169
WHO. (2015). WHO estimates of the global burden of foodborne diseases: Foodborne disease burden epidemiology reference group 2007–2015. World Health Organization (Chap. 5.1). https://apps.who.int/iris/bitstream/handle/10665/199350/?sequence=1
Yu, J. J., Wang, L., Wang, X. Y., Xu, J. P., Zhang, H. Y., & Yu, J. B. (2015, November). Research on the design of food safety supervision information system in Beijing. Proceedings of the 2015 5th International Conference on Computer Sciences and Automation Engineering, Sanya, China., 42, 618–624. https://doi.org/10.2991/iccsae-15.2016.115
Zheng, M. M., Zhang, S. S., Zhang, Y. D., & Hu, B. Z. (2021). Construct a food safety traceability system for people's health under the internet of things and big data. IEEE Access, 9, 70571–70583. https://doi.org/10.1109/Access.2021.3078536

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Research on the construction and functions of intelligent food safety supervision systems

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials And Methods

2.1. Data Collection

2.2. Data treatments

2.3. Data analysis

3. Results And Discussion

3.1. Distribution Analysis

3.2. Analysis of server and system performance

3.3. User groups analysis

3.4. Client construction analysis

3.4. The analysis of main functions

3.4.1. Food traceability function

3.4.2. Risk warning function

3.4.3. System management function

3.5. Characterized functions analysis

4. Problems Analysis

4.1. Non-unified data standards

4.2. Missing the multi-step dynamic risk early-warning model

4.3. Unreasonable and unprofessional function design

4.4. No systematic rules and regulations for system operation

5. Suggestions For Improvement

5.1. Promote the sharing of information between departments

5.2. Build the multi-step dynamic risk early-warning model

5.3. Optimize the function design

6. Conclusions

Declarations

References

Additional Declarations

Status:

Version 1