High-level reasoning while low-level actuation in cyber–physical systems: How efficient is it?
Journal of Industrial Information Integration, vol.51, 2026 (SCI-Expanded, Scopus)
- Publication Type: Article / Article
- Volume: 51
- Publication Date: 2026
- Doi Number: 10.1016/j.jii.2026.101090
- Journal Name: Journal of Industrial Information Integration
- Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC
- Keywords: Cyber-physical systems, BDI reasoning, Agent-oriented programming, Fuzzy logic integration, Worst-case execution time analysis, Embedded software engineering, Industrial informatics
- Dokuz Eylül University Affiliated: Yes
Abstract
The growing complexity of industrial information integration systems requires software technologies that support intelligent behaviour, real-time responsiveness, and efficient development. Despite the proliferation of programming languages and frameworks, there remains a limited amount of empirical evidence to guide engineers in selecting the most suitable tools for developing advanced industrial applications. This study addresses that gap by measuring and comparing worst-case execution time (WCET) and development time across six languages: C++, Java, Jade, Jason, and fuzzy Jason BDI with loosely and tightly coupled integration. These technologies represent a progression from procedural and object-oriented programming to agent-oriented frameworks that support symbolic and fuzzy reasoning. Instead of relying on broad or ambiguous notions such as paradigms or orientation, we adopt a developer-centred approach based on measurable outcomes. Our structured comparative analysis explores how increasing levels of abstraction and reasoning capabilities influence both the time required to develop applications and their runtime performance. By examining these dimensions, we reveal practical trade-offs among development effort and execution efficiency. Our findings demonstrate how different abstraction levels and reasoning mechanisms influence both system performance and engineering effort. These results provide practical insights for designing intelligent, agent-based systems that operate under real-time constraints and complex decision-making processes. The study contributes to the ongoing discourse on software selection in industrial informatisation by providing evidence-based guidance that aligns with integration efficiency, software maintainability, and system responsiveness. This work supports future research into the relationship between language features, development dynamics, and runtime behaviour in the context of industrial-oriented cyber–physical and smart manufacturing systems.