SYSE.9:11 - SoTA and Source Use
Derive specialist coordination from the Work that needs to be done. Keep the needed contribution, the System and its role classification or assignment, capability, authority, Method, performed Work, result and receiving decision separately recoverable. A list of professional titles leaves these relations to be established for the project.
| Source line | Retained contribution | Limit and guard |
|---|---|---|
| Grote et al. 2025 | A current positive Method derives organization-specific engineering-role bundles from required process contributions and stakeholder evidence; three industrial cases report clearer responsibilities and recognized gaps. | The conference study is limited to Advanced Systems Engineering organizations, uses judgment-laden workshops and one clustering technique, and does not merge kind, position, capability, assignment or Work. |
| Naikar et al. 2023/2024 | Complex human–AI design should include distributed teams, artifacts, networked technologies, communication, adaptation and self-organization rather than one human–machine task list. | This is a conceptual synthesis with an illustrative application, not validation of one complete Method; its institutional cases do not justify military or centralized-authority ontology. |
| Waterson et al. 2025 | Function allocation should include system interdependencies, joint operation, decision points, responsibility points, outcomes, authority and dynamic trust. | The framework and experiments are early and small; they establish neither universal responsibility allocation, AI moral agency, legal rules, nor a complete Work-design Method. |
| Becker et al. 2025 with the 2026 METR update, Agarwal et al. 2026, and Pradas Gomez et al. 2025 | AI Systems already perform bounded software and engineering-design Work; allocation needs task-specific capability, quality, provenance, integration and observation. | Software dominates the evidence, and the sources do not establish autonomous complete engineering, independent problem selection, authority transfer, role-holder replacement, or universal productivity. |
| INCOSE Competency Framework 2018 | Historical counterexample to title inference: role statements differ from job descriptions, one job can combine several roles, local tailoring and proficiency evidence matter. | It inherits a 2015 handbook, lifecycle/acquisition/defense framing and a fixed proficiency ladder. Its catalogue is neither current SoTA nor FPF ontology, assignment, Work or capability proof. |
| Team Topologies, second edition 2025 | Current author summary keeps cognitive load, platform grouping, whole-organization use and continuing adaptation visible as organization-design considerations. | The public page is not detailed chapter evidence or an independent effectiveness comparison. Software-team types and interaction modes are candidates, not universal engineering roles or DPF structure. |
Use an explicit epistemic status when field prevalence or causal effectiveness is not cheaply knowable. Expert judgment can guide a bounded engineering move; identify it as judgment rather than a population measurement. Reopen a contribution claim when a later engineering profile, role-derivation comparison, human–AI allocation study, actual project result, or changed capability makes the current request, holder, authority, acceptance, or DPF boundary wrong.