PaperlayerView exact revision on arXiv

Source-linked AI summary

Ubiquitous knowledge empowers the Smart Factory: The impacts of a Service-oriented Digital Twin on enterprises' performance

Francesco Longo, Letizia Nicoletti, Antonio Padovano

arXiv:2206.03268v1cs.CY

TL;DR

Industry 4.0 often emphasizes intelligence embedded in machines while overlooking human operators and the knowledge they need, a gap the paper frames as relevant to production and business performance. The authors propose a human-centric Industrial Internet pyramid and a Service-oriented Digital Twin to provide employees with ubiquitous knowledge access. Two company test-beds report benefits in time, costs, and process quality.

  • Problem

    The paper asks how human operators can be integrated with cyber-physical production systems to improve performance, noting that ubiquitous access to Digital Twin knowledge remains underinvestigated.

    Source: §1 · p. 4, Research question and study aims · p. 5

  • Method

    The authors propose an Industrial Internet pyramid and develop a Service-oriented Digital Twin that links the production system with employees to deliver manufacturing knowledge.

    Source: Research question and study aims · p. 5

  • Results

    Two complementary company test-beds report significant benefits in time, costs, and process quality from the proposed approach.

    Source: Research question and study aims · p. 5

  • Takeaways & Limitations

    The paper argues that a human-centric, knowledge-driven approach can support the Smart Factory transformation beyond technology adoption alone.

    Source: §5 · pp. 29–30

    As connected industrial assets and Service-oriented Digital Twins become more common, cyber attackers may target industrial IT systems to steal organizational knowledge.

    Source: §5 · p. 31

Abstract

from arXiv · show

While the Industry 4.0 is idolizing the potential of an artificial intelligence embedded into "things", it is neglecting the role of the human component, which is still indispensable in different manufacturing activities, such as a machine setup or maintenance operations. The present research study first proposes an Industrial Internet pyramid as emergent human-centric manufacturing paradigm within Industry 4.0 in which central is the role of a Ubiquitous Knowledge about the manufacturing system intuitively accessed and used by the manufacturing employees. Second, the prototype of a Service-oriented Digital Twin, which leverage on a flexible ontology-oriented knowledge structure and on augmented reality combined to a vocal interaction system for an intuitive knowledge retrieval and fruition, has been designed and developed to deliver this manufacturing knowledge. Two test-beds, complimentary for the problems in practice (the former on the maintenance-production interface in a large enterprise, the latter majorly focused in production and setups in a small and medium enterprise), show the significant benefits in terms of time, costs and process quality, thus validating the approach proposed. This research shows that a human-centric and knowledge-driven approach can drive the performance of Industry 4.0 initiatives and lead a Smart Factory towards its full potential.

1. Introduction

Industry 4.0 emphasizes intelligence embedded in physical systems, but the paper argues that manufacturing still depends on human workers and their access to useful information. The authors frame human-machine integration as important to production and business performance.

1. Introduction

  • Digital Twins manage knowledge from sensed data, but streamlined access and use by human operators is generally overlooked in their design.Workers remain necessary for operating and maintaining equipment and can adapt to workflow changes more quickly than robots, especially in mass production.Source: §1 · p. 3
  • Industry 4.0 efforts often focus on cloud computing and the Internet of Things while disregarding the human component and its impacts on production and business performance.The passage also notes that small and medium enterprises are hesitant to adopt such initiatives because of limited budgets and the need for quick returns on investment.Source: §1 · p. 4
  • Information quality, process uncertainty, and delivery speed affect decisions, while industrial workers need competencies to act alongside complex systems.The paper argues that the Smart Factory should move beyond a factory-of-things toward a human-centric manufacturing paradigm.Source: §1 · pp. 4–5

Research question and study aims

The paper asks how human operators can be integrated with cyber-physical production systems to improve production and business performance, given that ubiquitous knowledge access remains underinvestigated. It proposes a human-centric Industrial Internet pyramid and a Service-oriented Digital Twin, then evaluates the approach in two company test-beds.

Research question and study aims

  • The study asks how human operators can complement and integrate with cyber-physical production systems to improve production and business performance.It identifies access to and use of ubiquitous Digital Twin knowledge as an underinvestigated part of this question.Source: Research question and study aims · p. 5
  • The Industrial Internet pyramid positions a Service-oriented Digital Twin as a link delivering knowledge from the production system to manufacturing employees.The intended outcome is information symmetry between the cyber-physical system and employees.Source: Research question and study aims · p. 5
  • The prototype combines a flexible knowledge structure with augmented reality and vocal interaction for intuitive, ubiquitous access to manufacturing knowledge.The two complementary test-beds address maintenance-production in a large enterprise and production and setups in a small and medium enterprise; the paper reports benefits in time, costs, and process quality.Source: Research question and study aims · pp. 5–6

2. The Industrial Internet pyramid

The Industrial Internet pyramid extends the automation hierarchy by placing manufacturing employees and their ubiquitous access to system knowledge within the Smart Factory framework. A Service-oriented Digital Twin provides the knowledge link between employees and the cyber-physical production system.

2. The Industrial Internet pyramid

  • The conventional automation pyramid ranges from physical processes and smart objects to networked control systems and, at its apex, production and enterprise management systems.The passage distinguishes networked devices from cyber-physical production systems, which require integrated, context-dependent collaboration across production levels.Source: §2 · pp. 6–7
  • The Industrial Internet pyramid extends the automation hierarchy by making employees' ubiquitous access to manufacturing-system knowledge central.That knowledge covers system history, real-time state, and predicted future behavior, enabling employees to operate, maintain, and adjust the system.Source: §2 · p. 7
  • A Service-oriented Digital Twin delivers knowledge through accessible apps and services, linking the cyber-physical system with employees.The framework describes this as Level 4 of the Industrial Internet pyramid.Source: §2 · p. 7

3. Approach

The approach links the Industrial Internet and a Service-oriented Digital Twin to deliver manufacturing knowledge to employees through services, a configurable knowledge structure, and intuitive mobile access.

3.1. A service-oriented model for the Industrial Internet

  • The service-oriented model virtualizes manufacturing resources, packages CPPS data and documents as cloud-accessible services, and aims to help employees work more effectively, efficiently, and safely.Services are intended to support activities including maintenance and work-plan definition.Source: §3.1 · pp. 9, 10, 12
  • Managers receive services for monitoring resource status and potential faults, forecasting production scenarios, and receiving alerts about unexpected or unsafe CPPS behavior.The alert service is described as a proactive version of fault diagnosis and state monitoring.Source: §3.1 · pp. 10–11
  • Shop-floor operators use targeted fault diagnosis and augmented tutoring to investigate faults, access resource-specific guidance, and carry out tasks with support.Operators can act directly on resources; the paper notes remote action is restricted for safety reasons.Source: §3.1 · p. 11

3.2. A technological architecture of a Service-oriented Digital Twin

  • The distributed architecture connects the CPPS, its real-time digital reflection, networked services, and employee-facing remote terminal units, with third-party systems available for value-adding services.A machine-status request can return a JSON file.Source: §3.2 · pp. 12–14

3.2.1. A flexible ontology-based knowledge structure

  • The ontology-based structure represents assets with static attributes and configurable, sensor-linked dynamic attributes, allowing the knowledge structure to be adapted without programming expertise.An administrator can configure resource twins and connect them to physical resources using QR codes.Source: §3.2.1 · pp. 14–16

3.2.2. A technological solution for an advanced ubiquitous knowledge fruition

  • The Android operator app requests services through QR recognition, screen input, or voice, then interprets JSON responses into usable media; its speech features are reported at two seconds average delay and 95% reliability.The application uses Android Speech-To-Text and Text-To-Speech features.Source: §3.2.2 · pp. 16–17
  • Operator-accessible knowledge can include general machine information, 3D exploration of machine details, and the most worn component expected to need replacement soon.The cited example of the worn component is a safety switch.Source: §3.2.2 · p. 17
  • The application can also provide video guidance about a machine-element operation from the preceding manufacturing process and lists of the machine’s most critical past operations.Source: §3.2.2 · pp. 17–18

4. Application studies and discussion of the results

Two industrial case studies evaluate whether Service-oriented Digital Twin access to manufacturing knowledge improves performance. The large-enterprise study reports faster MWP revision and lower maintenance costs, while the carton-box study reports reduced setup and cycle times and statistically significant waste-rate reductions.

4. Application studies and discussion of the results

  • The application prototype was tested in large and small enterprises to assess quantitative effects on time, cost, and quality.The studies addressed maintenance-production coordination at the large enterprise and production setups at the small enterprise.Source: §4 · p. 18

4.1. The case of a large enterprise: Baker Hughes General Electric

  • MWP revision times differed significantly between the prototype-supported Group 2 and current-approach Group 1 for both machines.The reported p-values were almost null, and Group 2's mean was significantly lower for both machines.Source: §4.1 · p. 21, Table 1 · p. 21
  • The estimated maintenance savings cover operator labor for two machines; further benefits from avoiding spare-part waste and improving maintenance planning are described as potential.The authors also note the estimates do not cover all the large enterprise's machines and equipment.Source: §4.1 · p. 23
  • 16.96% total maintenance-cost savings were estimated for the fin tube machine, and 19.62% for the milling machine.For the fin tube machine, preventive and corrective maintenance savings were 14.84% and 22.71%; for the milling machine, preventive-maintenance cost fell 19.16% and corrective-maintenance time fell 21.24%.Source: §4.1 · p. 23, Table 2 · p. 24, Table 3 · p. 24

4.2. The case of a small and local enterprise producing carton packaging boxes

  • Across 100 production batches involving 10 operators, after smartphone application-prototype deployment, average batch cycle time decreased 7.57% (240.4 to 223.1 min) and setup time decreased 28.62% (63.8 to 46.6 min).Source: §4.2 · p. 25
  • Step-level improvements varied: setup time fell 44.5% in S1, 15.6% in S2, 17.1% in S3, and 24.9% in S4.The authors attribute the first-step improvement to tutoring and augmented assistance; they associate later-step improvements mainly with in-line diagnostics and condition monitoring.Source: §4.2 · pp. 25–26
  • The cycle-time and setup-time differences were statistically significant, and the production-waste-rate reduction was also reported as statistically significant.The authors state that the waste-rate reduction did not come at the expense of higher waste from lower setup and cycle times.Source: §4.2 · pp. 26–27, Table 4 · p. 28, Table 5 · p. 29

5. Conclusions & further research

The conclusions position ubiquitous knowledge and human–machine information symmetry as central to moving toward a Smart Factory. Two application studies report statistically significant performance improvements, while broader enterprise and supply-chain applications remain prospective.

5. Conclusions & further research

  • The Industrial Internet pyramid proposes that Smart Factory transformation requires reducing information asymmetry across technology, process, and people to achieve coupled human–machine interaction.The conclusions frame ubiquitous knowledge as the basis for this information symmetry.Source: §5 · p. 29
  • The application study reports statistically significant improvements in production and business performance and says the Digital Twin supports shop-floor employees at two companies.The conclusions identify the two-company application as evidence that performance benefits can be achieved beyond purchasing new technologies or machines.Source: §5 · p. 30
  • Service-oriented Digital Twins could support organization-wide change, customer services, and production rescheduling aligned with sales orders and expected production levels.The conclusions note the relevance of this coordination when maintenance is outsourced.Source: §5 · p. 30
  • Supply-chain use is a proposed extension, and wider connectivity makes cybersecurity and knowledge resilience important concerns.The authors suggest trusted data sharing technologies such as blockchain while warning that cyberattacks could expose organizational knowledge or affect physical systems.Source: §5 · p. 31
Loading 2206.03268v1…