The handbook of human–machine interaction – a human centered design approach

HMI – human–machine interaction? Is it not a far cry from 1970s favourite topic of human factors? Dethroned by HCI (human–computer interaction) or respectively by CHI (computer–human interaction), HMI had started to retreat into its shell in the 1980s. Now, it is back, and Guy A. Boy, the editor of the book, explains the rationale of ‘A human-centered design approach’ in the introduction. But it seems that the ‘machine’ he has in his mind is far more than the concept of most people associated with this word. To understand this, either one needs to read the introduction of the book and the last chapter by Erik Hollnagel about the diminishing relevance of HMI or just to study the so-called ‘Machine directive’ of the European Union. According to the latter, once I have been able to qualify a nuclear power plant as a ‘machine’ without fearing that legal people could argue against it. And the word ‘machine’ as a concept is more than ambiguous in any language. What it is depends on the degree of sophistication of the tools people use in a certain environment. While in some countries a bottle opener may be called a machine, in others, even complex equipment might be named ‘tool’ as if it were a screwdriver or a hammer. In the opinion of the editor of the book the entire HCI is part of HMI. Thus, the entity with which to interact may be a very complex system. Such systems once were planned as ‘automated’ systems, very often with the machine as the leading part and humans as operators. For example, initially the automation of nuclear power plants was planned to avoid human error by locking out human impact for the first 30 min of an incident. Later, it was recognised that in most catastrophes, the operator error plays a minor role compared to design errors or management errors – the end of the 30-min for safety story. The approach of this book reflects that notion, also in the title of the conclusion chapter ‘from automation to interaction design’. What I liked most while reading the book was that the content reflects the origins and personal background and history of its 38 authors. The editor, for example, is a university professor and director of the Human-Centered Design Institute (HCDi) at the Florida Institute of Technology, his PhD and habilitation theses were written in two French universities. Another author, Grote, has received her master’s degree in psychology from a technical university in Germany, her PhD from a US-university, and works at the ETH Zürich, Switzerland. A third author was born in Denmark, works presently in France, is a visiting professor in Norway, and a Professor Emeritus in Sweden. Organisational, engineering, cognitive, and social psychology, chronobiology, navigation modelling of complex Web applications – these and many more aspects, altogether an impressive list, can be found among the qualifications of the contributors. The book comprises 20 original chapters followed by a conclusion focusing on HMI. The order of the chapters resembles a design process; in Part I, methods for analysis are discussed in six chapters. The second part deals with design issues in seven chapters. Finally, in Part III, issues around the evaluation process are discussed. The chapters consider physical, cognitive, social and emotional aspects and deal with many key application domains such as aerospace, automotive, medicine and defence. Aerospace plays a major role in this book. No wonder, it is mainly about automation, but not in the sense of an automation created with the help of rigid machinery. The message it wants to convey is better characterised by the ‘Orchestra Model’ – all actors are highly skilled entities who would be able to ‘make music’ alone, but join a team under the leadership of one of them to play a beautiful symphony, a task that none of them can accomplish. I was delighted to encounter again the HABA/MABA approach (humans-are betterat/machines-are better-at) that has been used for some decades to explain the differences between the abilities of computers and humans in a new form, this time as an ‘un-Fitts list’, in this form telling why machines need people and why people create machines for. The list is commented in a chapter called ‘human–agent interaction’, very new for me although I must admit that I cannot remember how many times I had such interactions in the past few days. The chapter located physically in the middle of the book, forms in my opinion the core part of it by discussing the issue of authority and cooperation between humans and machines. The questions raised and answered here have been relevant since the first computer was assigned a job beyond being a calculating machine in the era of the mists of the computer Behaviour & Information Technology Vol. 30, No. 6, November–December 2011, 867–868