Is Now A Good Time?

Semmens, Rob; Martelaro, Nikolas; Kaveti, Pushyami; Stent, Simon; Ju, Wendy

doi:10.1145/3290605.3300867

Cited by 31 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The final TOR will be delivered when appropriate within the given time window (3) or at the end (with a proper lead time) when no suitable timing is found earlier (4). of transitions will be important to maintain the safety and convenience for multi-tasking drivers [12,22,27], which is a precondition for other activities in the car -for example, realizing concepts such as the "mobile office" [2].…”

Section: Discussion Limitations and Future Workmentioning

confidence: 99%

“…Such a system could become a valuable extension for in-vehicle interfaces, e.g., for multi-stage TORs [5]. Although the challenge of determining an appropriate timing to interrupt drivers has been proposed by empirical works both in the context of side activities [22] and transitions [28], we are not aware of any project investigating how such interfaces could be implemented. We have set up a driving simulation in Unity and modeled the scenario as an RL problem for our investigations.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The TOR Agent:Optimizing Driver Take-Over with Reinforcement Learning

Kuen

Schartmüller

Wintersberger

2021

13th International Conference on Automotive User Interfaces and Interactive Vehicular Applications

View full text Add to dashboard Cite

Figure 1: Learning process of the reinforcement learning agent. The agent observes a set of 10 coordinates (x/y, white boxes in the images) lying on the lane center of the upcoming road segment in exponentially increasing distance (i.e., the immediate future is represented in higher resolution). Given these observations, the agent can either issue TOR or postpone it to a later moment. By receiving the drivers' lateral performance after a TOR as reward signal, the agent learns which road trajectories are more suitable for TORs than others. Given the implemented artificial driver model, it would issue TOR in the straight segment (left), but postpone it in curvy sections (right).

show abstract

Section: Discussion Limitations and Future Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

The TOR Agent:Optimizing Driver Take-Over with Reinforcement Learning

Kuen

Schartmüller

Wintersberger

2021

13th International Conference on Automotive User Interfaces and Interactive Vehicular Applications

View full text Add to dashboard Cite

show abstract

“…The study utilises the data collection system for observational driving studies introduced by Martelaro and Ju [4] called "WoZ Way". This observation system provides a non-invasive, real-time, virtual, and remote presence to study naturalistic human interactions up close [5]. The WoZ Way system includes four go pro cameras, a quad-video multiplexer, a 9-DOF IMU, and a Mobileye 550 device [3].…”

Section: Apparatusmentioning

confidence: 99%

An Anthropological Study Designed to Understand the Essence of Intention Sharing Between Drivers and Passengers

Faramarzian

Pardo

Schroeter

et al. 2021

13th International Conference on Automotive User Interfaces and Interactive Vehicular Applications

Self Cite

View full text Add to dashboard Cite

This study focuses on human driving behaviour to identify key non-verbal cues which may inform a passenger of the driver's intentions. An anthropological inquiry, supported by live remote field observations and follow-up interviews, aims to understand a) the nuances and mechanisms, i.e. intention cues, through which human drivers consciously or unconsciously convey their driving intention, b) how passengers recognise and interpret those intention cues, and c) the role that the clarity, ambiguity or absence of these cues may play in passenger comfort or trust. Lastly, this research designs a live remote observation protocol to analyse the exchange of subtle intention cues between driver-passenger pairs during the driving task.

show abstract