An Analysis of Complexity Measures for Product Design and Development

This paper presents a new complexity metric for Object-Oriented (OO) design measurement to find at the design stage whether the classes become more complex, Moderate complex or less complex. The proposed metric is theoretically evaluated against the Weyuker’s properties as well as empirically evaluated against three open source software system. Furthermore, for validating the validity of new complexity metric, the paper has also presented the comparison of proposed metric with some well known complexity metrics like Weighted Method per Class (WMC) metric of Chidamber and Kemerer (CK), Class Complexity (CC) metric of Balasubramanian, and Complexity Metric for OO Design (CMOOD) metric of Rajnish and Bhattacherjee (RB) against the same three open source software system..Automated tool were used to generate the metric values and for analyzing the results, IBM SPSS software used. The results in this paper indicates that the new complexity metric is correlated well with existing complexity metrics and may be used as predictors of complexity of class.

Product complexity has been studied as an important factor to decrease the cost and time of the development process. With this purpose, prior research has included the development of design complexity metrics as a method to assess and decrease complexity. Recent studies have also focused on the comparison of complexity metrics for the particular case of medical devices development (MDD). However, the major issue relevant to MDD has not been addressed; the relationship between FDA regulations and the device complexity is not clarified. Therefore, to increase MDD safety and decrease the time to market, we must understand the regulatory decision process and rules. In this paper, we investigate the relation between different complexity metrics and FDA’s decision time using a sample of 100 hip replacement devices. Bayesian network learning is used to explore in detail local relationships between different variables, both complexity measures and product variables. This relationship was found significant for the first two clusters of the analysis. However, for a third cluster it is speculated that FDA decision time does not depend solely upon the degree of medical device complexity. Company or organization relevant variables could be playing a greater role than just complexity. Additional questions are drawn based on the results that must be investigated.

We propose a novel, lossless compression algorithm, based on the 2D Discrete Fast Fourier Transform, to approximate the Algorithmic (Kolmogorov) Complexity of Elementary Cellular Automata. Fast Fourier transforms are widely used in image compression but their lossy nature exclude them as viable candidates for Kolmogorov Complexity approximations. For the first time, we present a way to adapt fourier transforms for lossless image compression. The proposed method has a very strong Pearsons correlation to existing complexity metrics and we further establish its consistency as a complexity metric by confirming its measurements never exceed the complexity of nothingness and randomness (representing the lower and upper limits of complexity). Surprisingly, many of the other methods tested fail this simple sanity check. A final symmetry-based test also demonstrates our method’s superiority over existing lossless compression metrics. All complexity metrics tested, as well as the code used to generate and augment the original dataset, can be found in our github repository: ECA complexity metrics1.

This paper presents a novel integrated system of rapid product development for reducing the time and cost of product development. The system is composed of four building blocks — digital prototype, virtual prototype, physical prototype and rapid tooling manufacturing system. It can aid effectively in product design, analysis, prototype, mould, and manufacturing process development by integrating closely the various advanced manufacturing technologies which involve the 3D CAD, CAE, reverse engineering, rapid prototyping and rapid tooling. Furthermore, two actual examples are provided to illustrate the application of this integrated system. The results indicate that the system has a high potential to reduce further the cycle and cost of product development.

Regression modelling of software quality: empirical investigation

This chapter explores rapid descriptive sensory analysis methodologies (Check-All-That-Apply (CATA), Free Choice Profile (FCP), Flash Profile (FP), Pivot Profile, and Napping®) as efficient alternatives to traditional methods. These techniques enable the collection of reliable data using untrained panels or consumers, reducing time and costs in food product development. Their theoretical foundations, practical applications, and limitations are discussed: CATA is notable for its simplicity, using predefined lists of terms; FCP captures individual perceptions through free vocabulary; FP combines direct comparison with spontaneous description; Pivot Profile compares samples to a reference product; and Napping® maps similarities among products in a two-dimensional space. Practical aspects highlight the importance of selecting appropriate participants, designing balanced sessions, and employing specific statistical analyses. A case study with muffins illustrates how these techniques discriminate between formulations with different sugar content: CATA identified key attributes (“sweet” and “moist”), FP detected textural differences, and Napping® confirmed similarity-based groupings. In conclusion, these methodologies offer agility without compromising rigor, making them ideal for the food industry. Their selection depends on the study objectives, product complexity, and available resources. Future research could integrate them with advanced sensometric tools to enhance accuracy and product development.

The spatial organization of vegetation has been shown to be a strong indicator of ecological state in multiple ecosystems. In this study, we analyze the relationships between spatial complexity metrics in submerged aquatic vegetation (SAV) landscapes, and we explore the potential of satellite remote sensing to quantify these metrics in submerged environments. To do so, we estimated an array of complexity metrics over both simulated and real SAV landscapes of contrasted spatial organization. All these landscapes were artificially manipulated to (i) simulate remote sensing noise associated with the low signal‐to‐noise ratio (SNR) of aquatic environments and environmental noise generated by wind and waves, and (ii) reduce their spatial resolution from very high (2 m) to medium (30 m). Among these treatments, spatial resolution and low SNR (represented by sensor noise) had the strongest impacts on the perceived spatial complexity of the landscapes, while the impact of environmental noise was highly dependent on resolution. Although single metrics were deemed insufficient to characterize the spatial complexity of a landscape, a combination of informational complexity metrics such as the clumpy index, mean information gain, landscape shape index, and edge density provided a robust explanation of variation in the real and simulated datasets. These findings suggest that remote sensing has a strong potential for the ecological monitoring of SAV by contributing to establishing the link between SAV spatial structure and ecological status.

During the New Product Development (NPD) cycle, it is important to focus and integrate relevant information into the stages of product conceptualisation, design and development so that they can be used for product and process design effectively. In this paper, a Generic Product and Process Development System (GPDS), which integrates knowledge-based technologies of Rule-Based Expert System (RBES) and Case-Based Reasoning (CBR), is proposed to facilitate product and process development. Although RBES is adopted for sorting product design criteria, relevant past cases are retrieved by the CBR system during product and process design stages. In doing so, previous knowledge in terms of rules and past practices will be made use of during the product and process design stages, resulting in an enhancement of the efficiency of design and operations. By applying GPDS in Hasbro Fast East Limited, the new product design life cycle is greatly shortened with significant reduction in the total development cost.

Design for additive manufacturing (DfAM) creates opportunities for improved product design. Part consolidation (PC) is an important design opportunity in DfAM. The existing methodologies for part consolidation in DfAM are mostly based on expert judgment or heuristic rules. In addition, they only suggest the candidates for part consolidation. They do not assess the manufacturability of the consolidated parts. This paper proposes a two-level approach for part consolidation using complexity measures at the system level and part level. At the system level, the centrality score (a complex network measure) is used to identify parts with high potential for consolidation, and at the part level, a geometry-based complexity measure—modified complexity factor (MCF)—is used to assess the manufacturability. This two-step approach is expected to improve efficiency and effectiveness in DfAM since it uses minimal information at the system level and detailed assessment of manufacturability is done only for those parts where AM potential is high. Such an approach can be extremely useful to product designers at the early stages of design. Two case studies are presented to illustrate the effectiveness of the proposed approach.

Product development has become a focus of competition in many industries. Due to decreasing product life cycle, it is important to reduce the time and cost of product development. The product development process includes six main phases: product planning, concept development, system-level design, detailed design, testing and refinement, and production ramp-up [34]. Concept development process consists of four stages: identifying customer needs, establishing product specifications, and generating and selecting product concepts. Recent efforts have been made to improve the product development decisions [20], especially at the early stages of product development, since it has been recognized that nearly 75% of product life-cycle cost is committed by the end of concept development [26]. However, it is difficult to make product design decisions at the early development stages, because decision makers have to consider all life cycle issues [21, 22] with vague project information available.

Understanding source code is a cognitively demanding task, yet existing complexity metrics primarily rely on structural features and overlook actual human effort. This paper introduces a time-based approach to measuring code comprehension using large language models (LLMs). We investigate whether LLMs can estimate the number of seconds a junior developer would need to understand Java source files. Using a dataset of 371 files from the Apache Hive project, we evaluated five LLMs under a uniform prompt design. Human comprehension times, collected through a controlled study involving junior developers, were used as ground truth. Among the tested models, Gemini 2.0 Flash most closely approximated human estimates, while others exhibited systematic bias or output instability. Our findings show that while LLMs can produce interpretable, numeric predictions of comprehension time, achieving consistent accuracy requires prompt refinement and potential model fine-tuning. This work contributes a behavioral perspective to software complexity analysis, offering time-based estimations as an alternative to static metrics.

Unified Modeling Language (UML) has emerged as the software industry’s dominant modeling language. It is the de facto modeling language standard for specifying, visualizing, constructing, and documenting the components of software systems. Despite its prominence and status as the standard modeling language, UML has its critics. Opponents argue that it is complex and difficult to learn. Some question the rationale of having nine diagramming techniques in UML and the raison d’être of those nine techniques in UML. Others point out that UML lacks a comprehensive methodology to guide its users, which makes the language even more convoluted. A few studies on UML can be found in the literature. However, no study exists to provide a quantitative measure of UML complexity or to compare UML with other object-oriented techniques. In this research, we evaluate the complexity of UML using complexity metrics. The objective is to provide a reliable and accurate quantitative measure of UML complexity. A comparison of the complexity metrical values of UML with other object-oriented techniques was also carried out. Our findings suggest that each diagram in UML is not distinctly more complex than techniques in other modeling methods. But as a whole, UML is very complex–2-11 times more complex than other modeling methods.

Unified Modeling Language (UML) has emerged as the software industry’s dominant modeling language. It is the de facto modeling language standard for specifying, visualizing, constructing, and documenting the components of software systems. Despite its prominence and status as the standard modeling language, UML has its critics. Opponents argue that it is complex and difficult to learn. Some question the rationale of having nine diagramming techniques in UML and the raison d’être of those nine techniques in UML. Others point out that UML lacks a comprehensive methodology to guide its users, which makes the language even more convoluted. A few studies on UML can be found in the literature. However, no study exists to provide a quantitative measure of UML complexity or to compare UML with other object-oriented techniques. In this research, we evaluate the complexity of UML using complexity metrics. The objective is to provide a reliable and accurate quantitative measure of UML complexity. A comparison of the complexity metrical values of UML with other object-oriented techniques was also carried out. Our findings suggest that each diagram in UML is not distinctly more complex than techniques in other modeling methods. But as a whole, UML is very complex–2-11 times more complex than other modeling methods.

Technical analysis of intellectual property (IP) is conducted for the purpose of legal protection and product development. A brief review of the process of IP analysis and associated terminology is...

Software complexity is often hypothesized to be the enemy of software security. We performed statistical analysis on nine code complexity metrics from the JavaScript Engine in the Mozilla application framework to investigate if this hypothesis is true. Our initial results show that the nine complexity measures have weak correlation (ρ=0.30 at best) with security problems for Mozilla JavaScript Engine. The study should be replicated on more products with design and code-level metrics. It may be necessary to create new complexity metrics to embody the type of complexity that leads to security problems.