Driving Trust and Addressing Ethical Challenges in Transportation through Explainable AI
summary
Case study with Init in Karlsruhe builds a real-time dispatch support system using lightGBM trained on Init’s operations logs, surfacing per-prediction explanations with SHAP-style feature attributions and highlighted interface hovers that suggest short turns and other interventive actions. A small user evaluation (bachelor thesis) finds most dispatchers find the explanations helpful though some remain neutral; model is intentionally non‑automating, prioritizing human agency. Practical risks include over‑reliance, lost tacit skills, missing non‑digitized signals, and EU AI Act transparency requirements.
AI-written summary · about
chapters (11)
- 0:00Introduction and speaker background
- 1:12Research project overview and goals
- 3:28Role of dispatchers in public transport
- 8:11Initial AI tool feedback and transparency
- 10:51User involvement and explainable AI
- 13:02Key data sources for predictions
- 15:34Dispatcher evaluation of AI tool
- 17:27Human-AI collaboration benefits
- 21:42Risks and ethical challenges of XAI
- 24:42Ethical considerations and trust
- 25:47Q&A session
related talks
18mExtending Mobility - How AI Contributes towards the Future of Railway Transportation
16mBuilding Trust through Explainable AI
46mCan you trust your AI?
15mIntroduction to Explainable AI
39mThe Search for Transparency and Accountability in the Age of...
35mWhy Do You Need an Explainable Ai?
26mUsing Explainable AI to garner trust and overcome adoption inhibitors
01h 29mNavigating ML Explainability: From Theory to Implementation