Towards Measurement of Situation Awareness in Autonomous Vehicles
summary
A measurement technique called Daze issues real-time, in-situ event alerts (Waze-style) to probe driver situation awareness in both simulator and on-road studies. A 20-minute simulator course with 11 events and 12 participants yields 96% correct for visible+audible events, 83% for audible-only and 58% for visible-only; response times range 2.8–8.1 s and dispersion of location marks correlates with response time (r≈0.8). Comparison with SAGAT shows Daze is more ecologically valid but introduces interface-related misses; on-road use relies on pre-marked landmarks or a remote wizard.
AI-written summary · about
chapters (7)
- 0:00Introduction and project scope
- 1:20Measuring situation awareness
- 2:45Introducing the Daves tool
- 5:45Metrics and simulator study
- 9:00Comparing Daves with Sagat
- 11:59Real-world implementation
- 14:45Q&A and discussion
related talks
19mAssessing public perception of self-driving cars: the autonomous vehicle acceptance model
24mAdasa: A Conversational In-Vehicle Digital Assistant for Advanced Driver Assistance Features
17mGaze Augmentation in Egocentric Video Improves Awareness of Intention
45mTalking to your driverless car
18mThe Effect of Displaying System Confidence Information on the Usage of Autonomous Systems for ...
20mEffects of an Adaptive Modality Selection Algorithm for Navigation Systems
46mMelanie Bust | Compass UX Session 2025
11mInforming the Design of Privacy Awareness Mechanisms for Users and Bystanders...