Blockchain-Embedded Service-Level Agreement to Measure Trust in a Frugal Smart Factory Assembly Process

Integrating emerging Industry 4.0 technologies into smart factories has been widely discussed, particularly challenges regarding the practical use of a blockchain; one remaining challenge is the role of a blockchain beyond logistics and traceability, as well as its ability to support explicit trust measurement in real industrial environments. Existing studies often treat trust as a conceptual or cloud-oriented construction, without linking it to measurable production events. This study proposes a blockchain service-level agreement (SLA) to measure trust at an open-source frugal smart factory (SF). Trust is defined as a dynamic quantitative score derived from measurable process events, including estimated and response times, assembly correctness, and transaction outcomes; all of this is calculated through a smart contract implemented on a blockchain network. The approach is implemented in a tangram puzzle assembly process that integrates cyber-physical systems, edge computing, artificial intelligence, cloud computing, data analytics, cybersecurity, and the blockchain within a unified SF architecture. The framework was experimentally validated across four representative assembly scenarios: (i) the SF delivered the puzzle in time and was correctly assembled (<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>λ</mi><mi>s</mi></msub></semantics></math></inline-formula> = 0.1734), (ii) the puzzle was completed within tolerance time (<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>λ</mi><mi>s</mi></msub></semantics></math></inline-formula> = 0.0649), (iii) the puzzle was delivered on time and was incorrectly assembled (<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>λ</mi><mi>s</mi></msub></semantics></math></inline-formula> = 0.0005), and (iv) the puzzle was completed outside the tolerance time and was correctly assembled (<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>λ</mi><mi>s</mi></msub></semantics></math></inline-formula> = 4.91 × <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mn>10</mn><mrow><mo>−</mo><mn>5</mn></mrow></msup></semantics></math></inline-formula>); demonstrating that the model accurately estimates expected assembly times and updates trust without manual intervention during a physical manufacturing task, addressing the limitations of prior conceptual and cloud-based approaches. The main research contributions include an operational SLA-based trust model, the demonstration of the feasibility of applying blockchain-based SLAs in a physical SF environment, and evidence that a blockchain can be justified as a mechanism for managing and measuring trust in SF, rather than solely for traceability or logistics.