Skip to main content
SHIFT | Smart Hospital – Integrated Framework, Tools & SolutionsSHIFT | Smart Hospital – Integrated Framework, Tools & Solutions
Background Animation

Software Traceability: A Key Concept for LCNC Software Development Creating Software Models that Match Requirements from Patients, Practitioners, and Stakeholders

Knowledge database Structures & processes Data management & digitalisation Human Patient-centred approach C.3: Artificial intelligence-based software factory for MedTech applications

Software traceability ensures that stakeholder requirements—including those from patients and practitioners—are properly reflected in the software models built using Low-code/No-code (LCNC) tools.

Problem description, research question and relevance

In digital health, software development increasingly relies on collaboration between domain experts—such as healthcare practitioners, innovation managers, and administrators—and software teams. Low-Code/No-Code (LCNC) tools accelerate development by replacing traditional code with model-based software development.

However, ensuring that patient needs, software requirements, and standards are reflected in these models is challenging.

How can we ensure that what was requested is what gets implemented? Traceability is the key to this alignment. Yet, manual trace link discovery is time-consuming, error-prone, and unscalable given the diversity of LCNC tools and requirement formats. Thus, we ask:

  • How can we support trace link discovery between requirements and software models in LCNC-based software development?
Figure 1: OntoTrace Architecture. From Springer Nature Requirements Engineering: Foundation for Software Quality Conference [1]

Results and findings

Effectiveness: Subjects' average precision increased from 73.99% ± 15.77% without OntoTrace to 81.74% ± 15.49% with OntoTrace—an improvement of 7.75%. However, this difference was not statistically significant.

Efficiency: Subjects produced trace links at 1.44 ± 0.71 links/min without OntoTrace, versus 1.72 ± 0.71 links/min with it—an increase of 0.28 links/min (equivalent to 16.8 additional links/hour). This increase was statistically significant at 99% confidence.

Satisfaction: Using OntoTraceV2.0 improved average satisfaction across three measures:

  • Perceived Ease of Use (PEU): from 3.64 ± 0.86 to 4.11 ± 0.64
  • Perceived Usefulness (PU): from 3.27 ± 0.63 to 3.88 ± 0.69
  • Intention to Use (ITU): from 2.75 ± 0.94 to 3.38 ± 0.93

These increases (0.47, 0.61, and 0.63 points respectively) were statistically significant at 90%, 99%, and 90% confidence levels.

Figure 2: Quasi-experiment results’ distributions, having y-axis as the probability density. From Springer Nature Requirements Engineering: Foundation for Software Quality Conference [1]

Recommendations für practice

  • Bring traceability into digital health software development: Requirements and software models must be kept aligned—especially where regulatory and software quality is a must in domains as digital health. 
  • Reduce manual effort with automation tools. Automating trace link discovery increases efficiency and holds the potential to minimize human error. Using tools as OntoTrace may facilitate the adoption of good practices in software development teams in digital health.
  • Support Model and LCNC diversity: Future traceability tooling should be flexible across modelling formats and digital health software requirements. The integration with existing tools and other automation tools is crucial to allow traceability success.

 

Literature and other sources

[1] Mosquera, D., Ruiz, M., Pastor, O., Spielberger, J. (2023). Ontology-Based Automatic Reasoning and NLP for Tracing Software Requirements into Models with the OntoTrace Tool. In: Ferrari, A., Penzenstadler, B. (eds) Requirements Engineering: Foundation for Software Quality. REFSQ 2023. Lecture Notes in Computer Science, vol 13975. Springer, Cham. https://doi.org/10.1007/978-3-031-29786-1_10.

[2] Mosquera, D., Ruiz, M., Pastor, O., Spielberger, J., Fievet, L. (2022). OntoTrace: A Tool for Supporting Trace Generation in Software Development by Using Ontology-Based Automatic Reasoning. In: De Weerdt, J., Polyvyanyy, A. (eds) Intelligent Information Systems. CAiSE 2022. Lecture Notes in Business Information Processing, vol 452. Springer, Cham. https://doi.org/10.1007/978-3-031-07481-3_9

Quellenangabe

Back