Responsibility and Dependable Systems

This book brings together for the first time two important features of a computer system to be embedded in an organisational context: responsibility, meaning whether the computer system properly supports the organisational responsibilities that people are allocated; and dependability, which means whether it does so in a way that people can justifiably trust it to perform its operation correctly.

Organisational responsibility is examined from philosophical, social and practical viewpoints, which then leads to a workflow design which supports those responsibilities. The approach pays particular attention to looking at what happens when things go wrong and the adjustments and workarounds that people make to fulfil their responsibilities when their primary operational tasks are unachievable or frustrated, and how these alternative strategies can be incorporated into the workflow design. The application of the method to multi-organisational contexts is dealt with in some detail.

It is a truism that ‘human error’ is a common reason for failure in complex systems. These ‘errors’ sometimes lead to system failures because designers have failed to consider the normal human tendency to make mistakes. In other cases, ‘errors’ arise because of misunderstandings of who is responsible for what and how these responsibilities are to be discharged. Responsibility failures lead directly to actions that result in system failure.

Responsibility and Dependable Systems is the first book to examine the relationship between responsibility and system dependability. Both editors have backgrounds in computing and social science giving them a unique insight into how responsibility influences system dependability in different environments.

Divided into three parts, the first considers the philosophical and social aspects of responsibility, revisiting socially oriented system failures from a social, ethnographic perspective to tease out the complexities of the responsibilities that were implicated in the failure. Part Two uses a series of simple modelling notations to consider responsibility from a process and role-oriented perspective, using ideas from computer science, social science, management theory and engineering. The final section introduces new methods for analysing responsibility and mapping responsibilities within an organisation. The methods are illustrated by annotated models that can be adapted for use in both analysing and preventing failures of responsibility.

Systems engineers, computer scientists, social scientists, ergonomists and management researchers will all benefit from reading the book, and professionals and practitioners making organisational decisions concerning dependable human-computer systems, will also find it of value.