УДК 657

Estimating the efficiency of AI sorting mechanisms for dealing with the risks of excessive balance sheet inventory

Мараткызы Динара – студент магистратуры по программе “ Международная логистика(РУДН)” Казахского национального университета имени аль-Фараби.

Научный руководитель АдиловаНазданаДжемс-Уатовна – доктор PhD, старший преподаватель кафедры Бизнес-технологий Казахского национального университета имени аль-Фараби.

Abstract: In this study, the author evaluated the effectiveness of sorting mechanisms AI. An attempt is made to investigate how the use of AI-based applications can prevent the risk of accounting distortion caused by inventory management by comparing examples of enterprises using AI and not using AI. Based on this study, a noticeable difference was found in the efficiency of information storage along with the creditworthiness of the book value of inventories.

Аннотация: В данном исследовании автором проведена оценка эффективности сортировочных механизмов AI. Предпринята попытка исследовать, как использование приложений на основе AI может предотвратить риск искажения бухгалтерского учета, вызванного управлением запасами, путем сравнения примеров предприятий, использующих AI и не использующих AI. На основании данного исследования, выявлена заметная разница в эффективности хранения информации наряду с кредитоспособностью балансовой стоимости запасов.

Keywords: AI sorting mechanisms, book value of inventories, enterprises, deliveries, accounting.

Ключевые слова: сортировочные механизмы AI, балансовая стоимость запасов, предприятия, поставки, бухгалтерский учет.

Introduction

The central handling of the shipment of goods and services is known as supply chain management (SCM), and it encompasses all procedures that convert raw materials into finished commodities. The optimization of SCM is important for any type of parties as it allows to achieve the branding creditability, low operational costs, and high customers’ satisfaction. Hence, a significant number of resources both in industry and in research fields has been allocated to the advancement of the logistics sphere through minimizing the potential risks [1, p.3870-3871].

The biggest threat associated with supply chain management is considered to be economical, as its influence can be directly measured and quantified through the numerical values of financial reports. The data management focused on such risks and minimizing the costs has found several forms starting with the manual information analytics, quantitative models, and advancing with the most recent AI based approaches [2, p.2180]. The integration of the later model - neural networks - allowed suppliers to deal with the abundant number of variables and allowed to find optimal ways of product distributions. However, most of the developed models shift the goal from the conventional problem of financial distortions to the increase of supply chain process functionality. Due to this pattern, the issues like excess balance and accounting distortions in equity and liabilities, caused by the inventory management, can be overlooked.

Literature review

Supply chain management has been the subject of extensive discussion for a very long time. Organizations were always seeking ways to increase the efficacy and responsiveness of the logistics to achieve lower operational costs and soar its flexibility. Fisher, in 1997, has proposed the first methodological approach of supply chain administration that was primarily based on theoretical assessments and can be used generate the optimal supply chain alignment scheme. Further concepts were added such as more realistic assessment methods, enhanced variety of the logistics strategy, and addition of the competitive environment as an additional parameter [3, p.1]. The next step in the advancement of supply chain management maneuvers raised from the qualitative models, such as the demand and supply optimization-based networks, the reliability-based suppliers’ allocation, circular time reduction in logistics, and comparative model of the onshore and offshore manufacturers. In the contemporary world, the developments in the sphere of control of supply chains have faced the favorable integration with AI, IoT, and ML techniques [3, p.2].

When it comes to the term of inventory management, it is an integral part of business operation, as it is important to create an efficient system of inventory distribution, storage, and quantifying to achieve high customer services and low inventory costs. Several studies have been conducted on optimizing the inventory management protocols aimed to decrease the overall costs that were based on the assessments of the stochastic demand or lead time and on the utilization of the artificial intelligence as a coping mechanism [4, 257]. Despite that, accounting distortions regarded to the inventory still appear due to several possible reasons: abundance of the data, tracking & business planning difficulties, and costly operational budget [5, 229]. This paper will investigate the effect of the improper inventory quantifications on the accounting distortions by comparing the balance sheet values of the AI-based and conventional-based inventory managing organizations.

Methodology

A literature review and a case study comparison will serve as the two major methods used to gather the data for this project. The purpose of the literature research is to identify the essential characteristics and terms that will be further approximated in the case study and to provide an overall assessment of the industries using AI-based sorting mechanisms to manage the inventory component of logistics. It is intended to use the five-step primary data analysis technique developed by Denyer and Transfield. The five phases of this method include problem formulation, data collection, data evaluation, data analysis, and dissemination, which are intended to guide researchers through a structured and systematic approach to conducting research studies [6, 681-686].

The initial pilot research on the subject was based on a search procedure that is not restricted to any particular fields and focused mostly on developing the right working flow for subsequent phases. Since logistics, inventory, and accounting distortions are the three primary concepts in our study, this broad introductory approach is logical. Limiting the field will lead to biased and limited findings because so many different areas can have these kinds of attributes. The sources were investigated in different search strings, according to the protocol shown in the Table 1., to define the databases for the further use and generating the research questions for the study. The search string and scope were adjusted for each of the database separately to achieve the maximum coverage of related papers and exclude the non-relevant studies.

Table 1. Pilot studies’ search protocol.

Electronic Database

Search scope

Fields considered

Search string

Time period

Scopus

Title, keywords

All fields

“AI” AND “Artificial intelligence” AND “Supply chain” AND “Logistics” AND “Inventory” OR “AI” AND “Balance Sheet” OR “Balance Sheet” AND “Inventory”

2012-2022

ScienceDirect

Title, abstract, keywords

All fields

“AI” AND “Artificial intelligence” AND “Supply chain” AND “Logistics”

2012-2022

Web of Science

Title, abstract, keywords

All fields

“AI” AND “Artificial intelligence” AND “Supply chain” AND “Logistics” AND “Accounting”

2012-2022

IEEE Xplore

Title, abstract, keywords

All fields

“AI” AND “Artificial intelligence” AND “Supply chain” AND “Logistics”

2012-2022

Directory of Open Access Journals (DOAJ)

Title, abstract, keywords

All fields

“AI” AND “Artificial intelligence” AND “Supply chain” AND “Logistics” OR “Inventory” AND “AI”

2012-2022

JSTOR

Title, abstract

All fields

“AI” AND “Artificial intelligence” AND “Supply chain” AND “Logistics” OR “Inventory” AND “AI”

2012-2022

The results of the preliminary investigation defined the databases most suitable for the analysis in the topic of the effect of AI on the inventory management. The chosen engines are Scopus and ScienceDirect, since they provided the most appropriate range of credible literature for the topic despite having the unlimited field consideration. All the considered papers were written and published in English.

Results and Discussions

The first type of the analysis that was conducted is the distributional statistics of the found works. The total number of papers included for the analysis is 21 from the 4 main fields where the AI was implemented in the inventory management system. The 80% of the papers were obtained from the journal publications with remaining number being from the conference papers. The majority of the works were revealed after 2016 which can be explained with the fact that most of Industry 4.0 applications has only been recently included in the logistics system as illustrated in the graphical time-series distribution of works (Fig.1).

1

Figure 1. The time and field-based distribution of the chosen papers.

The following statistical analysis included the changes in the chosen 21 papers’ cases in terms of estimation of the book-to-market (BTM) ratio. Overall, 8 cases with equal split among 4 chosen fields were analyzed with results presented in Fig. 2. The case studies were based only on the enterprises that have implemented AI-based mechanisms into their inventory management.

2

Figure 2. Comparison of BMT values before and after AI implementation.

The appropriate book-to-market ratio, as determined by the basic concept, should be roughly 1. Lower or higher readings will indicate that assets or liabilities have been overstated or underestimated. The Fig. 2 shows the changes in BTM for all 4 sectors during the time before and after the adoption of AI-based technologies. Between samples taken before and after the application of AI, the changes were assessed using variance changes, and the significance of values were validated with the Z-test.

Overall, the empirical and descriptive statical analysis (descriptive statistical analysis, exploratory data analysis) have been performed. From the graphs and the statistical analysis, the changes in the BMT values have been dramatical for all the 4 sectors, proving that the implementation of AI into the inventory management field has indeed had an auspicious influence on reversing financial accounting distortions relevant to logistics.

The advantages of AI implication

Based on the conducted literature review, the most significant impact of AI on the role of inventory managers has been to reduce the amount of manual labor required to complete the task. AI has enabled the automation of many inventory management processes, allowing for more accurate and efficient operations. There are several aspects that shall be mentioned:

  • Artificial intelligence (AI)-based inventory control systems may be used to automatically track and monitor the quantity of products in stock, their location and expiration dates. This makes it possible to create judgments about inventory levels that are timelier and more precise.
  • AI may also be used to predict demand for different products, enabling better stocking choices that can assist to prevent overstocking and out-of-stock situations.
  • AI has also allowed for more sophisticated inventory management techniques to be employed. For example, AI can be used to track and analyze customer data, which can provide valuable insights into what customers are looking for and how they are likely to respond to different products or services, allowing businesses to make more informed decisions about their inventory management strategies.
  • AI can also be used to analyze large amounts of data to identify areas of inefficiency or wastage in the inventory management process, allowing for more efficient use of resources.
  • AI creates a shift to the strategic planning. The increased efficiency and accuracy of the inventory management process has allowed inventory managers to focus more of their time on strategic activities. This, in turn, has allowed organizations to take a more active role in the development of new inventory strategies, as well as engaging in activities such as market research and customer relationship management.

The modifications made to inventory management as a result of AI have also reduced accounting distortions. A company's inventory levels may be analyzed using AI technology, which can also spot any irregularities or data inaccuracies. By doing so, it is possible to prevent any potential financial statement distortions and guarantee that the inventory is recorded appropriately. Finally, AI has made it simpler for firms to identify any fraudulent inventory management-related activity. Artificial intelligence (AI) tools may identify any unusual trends in the data and notify them of any potential anomalies. This lessens the possibility of accounting distortions brought on by illegal transactions.

The disadvantages of the AI implication

The future of the labor market may be impacted by the integration of AI into the inventory management process. There will likely be fewer employment available in this industry as AI becomes more pervasive and powerful and more manual labor tasks within the inventory management process are expected to be replaced by AI-based solutions. As there are already many individuals working in manual labor tasks in the inventory management process, this might have a substantial negative influence on the job market.

Additionally, upcoming changes in the labor force system may cause dramatic changes for the existing norms of financial statement analysis. Implementation of additional factors in estimating the inventory, integrating the AI-based machines as a part of equipment may results in challenges such as calculating the appropriate amortization, depreciation rates, and goodwill.

Derived existing gaps

Conducted literature review, assisted to identify that the current research on AI implementation in inventory management focuses primarily on the technical aspects of implementation and the potential benefits of AI-driven inventory optimization, leaving the gaps in the areas of organizational and cultural implications of AI integration, such as the potential for resistance to change, the need for organizational buy-in, and the impact of AI on employee roles.

Conclusion

It was discovered that the supply chain management system is often associated with economic risk, which can be quantified and measured by financial reports. Over time, data management solutions have been developed to minimize costs, and particular attention has been paid to the neural networks that enable suppliers to manage a large number of variables and find the most efficient product distribution. The topical relevance of this study was proved based on the primary data collection, showing that while the modern analytical models shifted the focus away from traditional financial concerns, it created an overlooked gap in the topic of inventory distortions.

Based on the conducted literature review and assessed case-studies, the introduction of AI into the inventory management space has had a significant impact on the role of inventory managers. AI has enabled the automation of many processes, resulting in increased efficiency and accuracy, as well as allowing inventory managers to focus more of their time on strategic activities. Furthermore, they have increased the accuracy of book-to-market inventory values and assisted businesses in streamlining their supply chain management procedures in the way of minimizing accounting distortions. However, there are potential implications for the job market, as more manual labor roles may be replaced by AI-based systems in the future, resulting in a reduction in the number of jobs available in this area. The changes in the employment system may create vastly resource-draining transformation for the in the system of financial statement calculations as AI tailored factors and alterations will be required in several fields (e.g., depreciation, R&D expenses).

The further studies in the field of the AI, inventory management, and the accounting can be focused on the identified gaps in the existing literature on the topics: the influence of AI on employee responsibilities, performance, and the requirement for corporate buy-in.

References

  1. Singh D. Inventory Management in Supply Chain / D. Singh, A. Verma // Materials Today: Proceedings. – 2018. – Т. 5, № 2. – С. 3867-3872.
  2. Supply chain risk management and artificial intelligence: state of the art and future research directions / G. Baryannisa, S. Validi, S. Dani, G. Antoniou // International Journal of Production Research. – 2019. – Т. 57, № 7. – С. 2179–2202.
  3. Mahdavi M. Efficient vs. Responsive supplier selection for functional and innovative products: A quantitative study / M. Mahdavi, M. Mahmoudzadeh, T. L. Olsen // International Journal of Production Economics. – 2023. – Т. 255, № 108656. – С. 1-13.
  4. Praveen U. Inventory management and cost reduction of supply chain processes using AI based time-series forecasting and ANN modeling / U. Praveen, G. Farnaz, G. Hatim // Procedia Manufacturing. – 2019. – Т. 38 – С. 256-263.
  5. Osman B. M. A study on role of artificial intelligence to improve inventory management system / B. M. Osman, S. Alinkeel, D. Bhavshar // International Research Journal of Modernization in Engineering Technology and Science. – 2022. – Т. 4, № 3. – С. 226-233. – ISSN 2582-5208
  6. Denyer D. Producing a systematic review. / D. Denyer // The SAGE Handbook of Organizational Research Methods / D. A. Buchanan, A. Bryman. – UK : Sage Publications Ltd, 2009. – 39. – С. 671–689.