Goals Achieved and Missed
This page is based on the binding target agreement (Zielvereinbarung) (project-management practice project SS 2026, signed by client Lily Dausch, Bundeswehrkrankenhaus Berlin — represented by head of nursing Dagmar Klauck —, course lead Prof. Dr. Heike Wiesner, and Kâan Turan). An earlier version of this page documented the mid-project state; it was updated to the final state after the final presentation (14.07.2026).
Overall result: mission accomplished
All eight goals of the target agreement were achieved:
| Priority | Achieved |
|---|---|
| MUST | 2 / 2 ✓ |
| SHOULD | 2 / 2 ✓ |
| COULD | 4 / 4 ✓ |
The eight goals in detail
MUST goals
| No. | Goal | Success criterion | Result |
|---|---|---|---|
| 1 | Introduction of different patient profiles (age, gender, origin, skin color, body type, mobility) | Integration of 5 additional patient models | ✅ Fulfilled: 6 diverse profiles (A–F) in DB, frontend, and Unity scene — child, non-binary person, middle-aged woman, senior with dementia, obese patient with amputation, patient with language barrier / homelessness (Patient Models) |
| 2 | Assessment for improving accessibility (e.g. motion sickness, color blindness) | Status quo + written recommendations (approx. ½–1 page) | ✅ Fulfilled: written assessment by Erik Wlochal — status quo plus prioritized recommendations for future development (Assessment Barrierefreiheit, DOCX); color-vision modes implemented in the frontend (protanopia, deuteranopia, tritanopia, achromatopsia) |
SHOULD goals
| No. | Goal | Success criterion | Result |
|---|---|---|---|
| 3 | VR patient file as preparation for further (diversity-oriented) error sources | Clickable patient file exists; completeness flexible | ✅ Interactive clipboard in VR (PatientFileUI): multi-page file per patient, with file errors and decision dialogs (interpreter, missing address, allergy) |
| 4 | Fixing technical errors and improving stability | At least 50 % of the improvements raised by Kâan resolved | ✅ 100% Goals Fulfilled: Incl. PostgreSQL migration, native UUIDs, Zod validation of all endpoints, startup config validation, error logging, WebSocket layer, patient-bed decoupling in Unity |
COULD goals
| No. | Goal | Success criterion | Result |
|---|---|---|---|
| 5 | Comprehensive manual for hosts and users | First draft / concept proposals sufficient | ✅ Exceeded (two complete manuals instead of a first draft): detailed German user manual for hosts and non-technical users by Semi Kaan Baskaya (Manual Room of Horror, DOCX) plus a VR-focused manual for trainers and lecturers by Jacob Gotter (Nutzerhandbuch Room of Horror VR, PDF); additionally setup and workflow guides (Local Setup, Game Start), FAQ page in the frontend, tutorial video (room-of-horror-assets/Room of Horror Tutorial.mp4), onboarding tour in the dashboard |
| 6 | Diversity-oriented expansion of existing error sources | 1 new error source added | ✅ Exceeded: 13 new errors (IDs 19–31), all diversity-related — incl. child bed, bariatric bed, dementia sensor mat, interpreter decision (Error Catalog) |
| 7 | Weighting of errors by severity of medical consequences + revision of the results dashboard | Revision of the weighting | ✅ Weighted error catalog (weight 1–5, criticality, categories), rule-based evaluation engine, new results dashboard (EvaluationPanel with category breakdown and trainer feedback) |
| 8 | Evaluation / prototypical integration of AI-supported functions with diversity relevance | Short feasibility study outlining application areas | ✅ Feasibility study by Jan Busson — Room of Horror 3.0: KI-Machbarkeitsstudie (PDF): three solution approaches (AI-generated audio tracks, controlled speech interaction, dynamic virtual patient) with Unity roadmaps, diversity application areas, and open questions for future teams; AI patient conversations deliberately not implemented in 2.0 |
Deliberately excluded (per target agreement, section “Delineation”)
- Complete redevelopment of the application
- Production-ready launch / app-store release
- Photorealistic rendering of the scenarios
Additionally de-scoped during the project (stability before feature count): implemented AI patient conversations, multilingualism, hint system, subtitles — see Requirements Catalog.
Reflection
What went well:
- Early and regular alignment with Lily Dausch kept the requirements realistic
- Focusing on stability instead of feature count proved itself
- Iterative approach with weekly feedback rounds
What we would do differently:
- Fix requirements in writing earlier (target agreement only in week 9)
- Consider accessibility from the start instead of as a downstream task
Information sources
- Target agreement “Room of Horror – Weiterentwicklung der VR-Trainingsanwendung” (SS 2026): goal list, success criteria, priorities, delineation, stakeholders
- Final presentation “Room of Horror 2.0” (14.07.2026, PDF): overall result “100 % Mission erfüllt” (mission accomplished)
- Deliverable documents: Assessment Barrierefreiheit (Erik Wlochal), Manual Room of Horror (Semi Kaan Baskaya), Nutzerhandbuch Room of Horror VR (Jacob Gotter), KI-Machbarkeitsstudie (Jan Busson), Projektdokumentation (Jacob Gotter et al.)
- Repository code as implementation evidence (incl.
ColorVisionControl.tsx,PatientFileUI.cs,database/General.ts,evaluation/*,websocket/*,components/results/*)
Last build: 19 Aug 2026, 05:20+00:00