Evaluation

We validated the results of the comparative analysis through an expert questionnaire survey complemented by targeted follow-up interactions. Experts evaluated our findings in terms of completeness, correctness, and alignment with established practice.

Overview

The evaluation focuses on validating the results reported for RQ1-RQ3. The survey includes Likert-type questions and structured qualitative feedback to support the interpretation of disagreements and to refine the analysis where needed.

Study design

Questionnaire structure

  • Self-assessed expertise per formalism (1-5 scale)
  • Likert agreement questions (1-5 scale)
  • Mandatory open-text justification for ratings ≤ 3

Follow-up interactions

  • Targeted to deepen/clarify critical feedback
  • Selection based on expertise and response insights
  • Synchronous/asynchronous depending on availability
Replication note: the full questionnaire and anonymized responses are provided in the replication package. You can add links below in the “Materials & data” section.

Participants

83

Invitations sent

29

Complete responses

34.94%

Response rate

Responses were collected over four weeks (January 2026). The respondent group included a mix of academic and industry profiles (e.g., PhD students, postdocs, researchers, professors, and practitioners), with an average experience of approximately 8 years in robotic software engineering. Follow-up interactions were conducted in February 2026.

Note: We collected informed consent from all respondents prior to participation, ensuring that they understood the purpose of the study and their rights regarding data usage.

Results summary

RQ1 - Control structures and mission concepts

Experts evaluated the mapping between formalism primitives and the control structures / mission concepts. Feedback was used to confirm or refine the interpretation of constructs and the consistency of the mapping.

RQ2 - Expressivity, peculiarities, and limitations

Most respondents agreed with the expressed support levels and the identified strengths/weaknesses. Disagreements mainly concerned concerns that can be addressed through implementation-level extensions (e.g., time-dependent behavior, state saving & resuming, explicit waiting), and were discussed and integrated into the final observations.

RQ3 - Tool support

Experts validated the tool taxonomy (scope and capabilities), reported their familiarity with the tools, and suggested missing tools or overlooked aspects. Feedback informed refinements of the tool tables and discussion.

Materials & data

Questionnaire: Link to questionnaire

Responses: Open the spreadsheet in a new tab

Follow-up transcripts: par:5 | par:10 | par:17 | par:20

Note: transcripts are anonymized and shared for replication purposes.

Responses preview

Anonymized responses spreadsheet (read-only).