Book Review: Automobile Automation: Distributed Cognition on the Road

Automobile Automation: Distributed Cognition on the Road is a timely, academic exploration of the human factors issues associated with in-vehicle automation. The authors discuss the relative aspects of traditional driver-vehicle interaction and how roles and responsibilities change with the introduction of automated systems. A systems framework helps untangle some issues associated with distributed cognition for varying levels of automated vehicle systems.

A theoretical concept, distributed cognition describes cognition as being a holistic sum of all potential actors during a task. In the case of driving, cognition may be distributed among the human, various automated or semi-automated features of the vehicle, and additional resources such as the driving environment and infrastructure. In Chapter 3, the authors break down the potential allocation of traditional information processing functions (“Monitor, Anticipate, Detect, Recognize, Decide, Select, and Respond”) to the driver versus the automation when using various types of vehicle technology. For example, power steering has most if not all functions allocated to the driver (with the potential for automated aid in the “Respond” stage), but collision avoidance systems allocate all functions to the automation (with the driver monitoring the automation as opposed to the task).

The book admirably describes tools that practitioners in the automobiles and automation field might use to explore driver perceptions and behavior as they relate to automated vehicles. Specific approaches that may be useful include:

the collection and analysis of concurrent and retrospective verbal protocols to assess driver perceptions of automated systems

the application of additional task analysis methods including operational sequence diagrams (OSDs), critical decision method (CDM), and link analyses to interpret driver behaviors

the use of traditional social network diagrams to quantify the interactions between the driver and automated vehicle components

In the context of the distributed cognition framework, these tools are intended to help readers empirically identify which roles should be allocated to the human and which to the vehicle or other agents and, furthermore, when there is potential need to reallocate roles or functions.

I particularly liked the last chapter (“Summary of Findings and Research Approach”). The authors made a point to pull together the general theory and overview of the usable tools presented throughout the book, a summary of all findings, and arguably most importantly, insightful suggestions for future research avenues. The challenge, as tackled by the authors, is to appropriately define the allocation of functions of the driving task and to do so in such a way as to create a safer environment that, surprisingly, may not always include allocating all tasks to automation. Specifically, the authors refer to “automation surprises,” meaning unexpected behavior of the automation, or potential cases in which the driver has become so far out of the “loop” that he or she is unable to resume the driving task should it become necessary.

This book is intended for a broad audience, with the tone most often skewing toward the academic. It would be especially useful reading for newcomers to automobile-automation research as it gives a succinct overview of the current issues, related research, and appropriate tools to study driver perceptions and behaviors. This book could be used as a reference guide for researchers but would also be a great introduction to human factors for automobile automation for nonacademics or researchers in other fields.