This website contains the official supporting materials to the paper “Integration Complexity,” submitted to the 36th Annual INCOSE International Symposium (IS2026). The symposium was held June 13-18 at the Pacifico Yokohama Conference Center in Yokohama, Japan.
The literature contains vast empirical evidence that software development has become more complex over recent years as engineering teams strive to develop more modular, feature-rich, and higher-performing software components. Consequently, accurately quantifying and tracking the evolving complexity of software components has become increasingly critical. While the software engineering community has conducted considerable research on developing complexity metrics to assess the internal complexity of software components, there is no evidence in the literature that indicates steps towards creating a metric to evaluate the complexity required to integrate two software components.
The paper outlines a proposal for a new software complexity metric, called Integration Complexity (NTC), a simple, quantifiable measure for evaluating such complexity. While the work presented focuses on application-level integration and is consequently oriented towards interfaces based on application programming interface contracts, we expect that future NTC proposals will include additional functional proposals for other integration types, such as data-level and message-based integration.
The motivation for proposing the complexity in rooted in: (1) the ubiquity of software across modern engineered complex systems, (2) the importance of software interfaces as the foundation for component communication and as a prerequisite for system integration, (3) software engineering and systems engineering practitioners share the responsibility for defining, developing, and maintaining software interfaces, (4) the imperative need to assess and track the evolution of complexity in software interfaces, (5) the drive to find a solution to facilitate the assessment, tracking, alerting, reporting, and visualization of changes in software interfaces to assist globally distributed, multidisciplinary teams working on software-intensive systems, (6) the drive to formulate a solution as a software complexity metric, thereby accelerating its adoption across existing processes and tools, and (7) the lack of empirical evidence in the literature on a software complexity metric, such as NTC.
The plots SDG001 to SDG006 demonstrate initial exploratory results. They were generated in Wolfram Mathematica (Version 14.3) using synthetic datasets and depict the Integration Complexity Plane (ICP) under different conditions. The Synthetic Data Generator, the internal Java domain-specific language (DSL) developed to simplify configuration, and the datasets used in the study are publicly available to any researcher interested in reproducing the results.
We anticipate that NTC can provide the foundational building blocks for a globally distributed system1 to help manage the complexity of component interfaces worldwide.
Author and Contact Information
Citation and Paper Availability
Icn. 360505011113033603