Will drivers cognitively disengage from previous tasks following scheduled takeovers? Identifying the influencing factors of drivers' attention residue

In Level 3 automated driving, drivers can engage in non-driving-related tasks (NDRTs) but must resume control when required, known as takeovers. While prior research has predominantly focused on drivers' overt responses to task interruptions during takeovers, the present study examines an underlying cognitive challenge. In particular, drivers may experience attention residue, where cognitive engagement with prior NDRTs persists after switching to manual driving, potentially impairing post-takeover performance. This study employs a mixed-methods approach to investigate factors influencing attention residue during scheduled takeovers. Qualitative interviews identified nine potential factors, six of which were tested through a quantitative experiment under varying driving demands (low vs. high). Generalized linear mixed models confirmed situational motivation, mind wandering tendency while driving, and driving demand as robust predictors of self-reported attention residue. Furthermore, the findings suggest implications for developing targeted interventions to mitigate attention residue in takeover contexts and provide valuable insights for informing future research.