Production logistics digital twins: Research profiling, application, challenges and opportunities

Abstract

In the era of Industry 4.0, Production Logistic Digital Twins (PLDTs) have garnered remarkable attention from both academic and industrial communities. This is evident from the growing number of research publications on PLDTs in international scientific journals and conferences. However, given the diversity and complexity of production logistics activities, there is a pressing need for systematic literature review to chart past research and identify potential directions for future endeavors. Therefore, this study primarily focuses on the application of Digital Twins (DTs) in Production Logistics (PL). Firstly, an analysis of PLDTs research profiling is carried out based on general trends, keywords, application scenarios, and basic functions. Secondly, the functional characteristics of PLDTs are examined while summarizing their advantages and limitations across various application scenarios such as transportation, packaging, warehousing, material distribution, and information processing. And the roles played by smart technologies such as Internet of Things (IoT) in PLDTs system are discussed. Finally, possible challenges and future directions of PLDTs in industrial application are presented, accompanied by appropriate classification and extensive recommendations.

Introduction

Manufacturing enterprises are experiencing subtle changes under the impact of Industry 4.0, particularly in the field of Production Logistics (PL). PL is committed to the direction of automation, digitalization and wisdom with thinking, perception, learning, reasoning and autonomous decision-making capabilities [1]. As an essential link between shop-floor supply and production processes, PL accounts for approximately 95% of the entire production life cycle [2]. This shows that effective and reasonable PL management plays a momentous role in augmenting the competitiveness and economic efficacy of enterprises, especially for the intelligent development and digital transformation of today's enterprises.

The initial focus of the PL management approach was on scientifically planning, allocating, and controlling material flow within production processes. However, with the emergence of Industry 4.0, the emphasis of PL management has shifted towards information flow management, resulting in an increased demand for flexibility, agility, and consistency of real and virtual interactions in the Production Logistics System (PLS) [3]. This poses various challenges for researchers and practitioners alike. For instance, (i) how to effectively connect the virtual and physical worlds while realizing seamless integration and real-time interaction in the complex technical landscape [1]. (ii) How to systematically integrate heterogeneous systems and aggregated data platforms under big data environments characterized by massive volume and heterogeneity [4]. And (iii) how to optimally utilize economical computational resources to accomplish the most precise synchronous control of PL in complex environments featuring varying levels of dynamic disturbances [5].

Aligned with Industry 4.0, the concept of Digital Twins (DTs) has been extensively employed in manufacturing research since its inception. DTs is aimed at representing and optimizing physical objects through virtual models driven by a combination of data and models. Recently, the terms Logistics Digital Twin (LDT) and PLDTs have appeared in many works, with the goal of improving the performance of logistics. For example, Piancastelli et al. [6] derived the architecture of LDT and pointed out that LDT helps management to make decisions, evaluate and reduce the risk of Production Logistics Activities (PLAs) scenarios, etc. Thürer et al. [7] proposed a new architecture for PLDTs. Kaiblinger et al. [1] summarized the common definitions of DTs in the field of PL. More studies have discussed that DTs brings new concepts, models and ideas to the intelligent operation of logistics with its characteristics of virtual simulation, evaluation, prediction and autonomous decision-making [8]. It is also shown that DTs is an effective way to realize the integration, interaction and intelligent interconnection of production and logistics processes in the virtual world and the physical world [4].

In recent years, there has been a proliferation of reviews exploring PLDTs, with each offering unique insights. (i) In terms of research interests, Pawlewski et al. [8] described the research implications of using DTs to optimize intralogistics processes and named it Digital Twin Lean Intralogistics. The research results have indicated a growing interest in the terms “digital twins” and “intralogistics”. (ii) As to the virtualization integration of PLS, Fottner et al. [9] discussed recent research advances in autonomous systems in intralogistics. And the importance of DTs, modeling and simulation techniques for the virtualization of the entire intralogistics system was emphasized. Kosacka-Olejnik et al. [10] addressed the question of how the DTs concept can support internal transport systems. Explained that DTs supports internal transport systems by establishing dynamic links and correspondences with real objects and internal transport processes. (iii) As for the application scope of DTs, Zafarzadeh et al. [11] classified PLAs into 3 categories. The share of 10 groups of enabling technologies in PLs, such as DTs, was systematically reviewed and evaluated. It was also shown that there are applications of DTs in PLAs such as tracking and location, material distribution and warehousing. Most recently, Kaiblinger et al. [1] discussed the common definitions, features and functions of DTs in the field of PL. The current state of development and implications of the latest implementations were outlined, and 20 application cases were evaluated to identify current research gaps.

As previously mentioned, researchers have emphasized the role of DTs as a crucial factor for virtualizing PL processes in their analysis of PLDTs. This is consistent with the urgent need for information-physical fusion technology during the digital transformation of today's manufacturing industry. To further describe the role of DTs in specific PLAs, this study focuses on the latest applications of DTs in PL and supplements the discussion of common application scenarios, methods or theories, and related intelligent technologies of existing PLDTs from the perspective of functional characteristics. Firstly, the common application scenarios of DTs in PL are discussed based on keyword analysis, including transportation, packaging, warehousing, material distribution and information processing. And the distribution of decision support, simulation, planning, monitoring, evaluation, tracking and positioning, predication and design in each logistics scenario are analyzed. Then, more detailed analysis of the methodologies, functions, and application validation methods of DTs in specific PL activities are analyzed. The roles played by intelligent technologies such as Internet of Things (IoT), big data, and Cloud Computing (CC) in PLDTs system are summarized and discussed. Secondly, the application advantages and challenges of DTs in PL are discussed. Finally, the possible future directions of PLDTs are described from the perspective of industrial applications. This study is a complement to the above research work, providing an appropriate classification method for existing theoretical studies and industrial applications of PLDTs.

The remainder of the paper is structured as follows: Section 2 describes the research profiling of PLDTs. Section 3 provides a detailed analysis of the role of DTs in PLAs. Section 4 describes the role that intelligent technologies play in PLDTs system. The advantages and challenges of existing research methods are summarized in Section 5. Furthermore, Section 6 describes the possible future directions of PLDTs in industrial applications. Finally, to conclude this paper.