Coordinating multi‐level cognitive assistance and incorporating dynamic confidence information in driver‐vehicle interfaces

Automation technologies have now been widely introduced into many aspects of human–machine collaboration. Most automated systems, however, are still far from perfect in performing assistive duties for their operators. This situation can lead to chaotic assistance, which can easily cause cognitive confusion and performance degradation among operators. This article discusses how to coordinate multiple types of assistance that have different levels of cognitive intrusiveness. To improve the cognitive adaptability of operators, dynamic confidence information is presented on the interactive interface to indicate the likelihood of the correctness of such assistance. Multiple types of assistance are then presented at different moments to match human‐cognitive‐processing stages and eventually to attain a coordinated status. In a car‐following driving experiment, the drivers' responses to multiple types of assistance with dynamic confidence information were investigated. The coordinated assistance successfully enhanced the drivers' driving performance and the human–machine relationship. This method can be further generalized to improve human–machine collaboration in many industrial automation systems that still require human operators. © 2012 Wiley Periodicals, Inc.