Remote Cooperation:
CSCW Issues for Mobile and Tele-Workers

Alan Dix and Russell Beale (eds.)


Introduction

Alan Dix and Russell Beale


Contents:

So Near Yet So Far
The Structure of This Book
Framing the Problem
Future Directions and Further Information
Notes

top | intro. contents | notes | book home page

So Near Yet So Far

We in a global village. You can live in Andorra, yet telephone your friends in Australia, log in to a computer in Argentina and watch live American football. The only things which separate us are our time-zones. Once maps of the world had "here be dragons" at their edges; now we can see images of the Earth from space over the Internet (note 1). We are living in a shrinking world, but it is not shrinking uniformly. There may be few square kilometres of the populated world without a television aerial, but there are many without a telephone.

Our definitions of distance are changing. Ten thousand miles is no barrier to communication, but ten minutes' walk may be. Our friends in Australia may hear more from us than our Auntie in the next street. Today in many conurbations the ten minute walk in the street at night may be more perilous and alien than ever were those indistinct margins of the medieval map.

The word 'remote' has correspondingly taken on new meanings. Once a remote country was far away; now it merely means far away from a road. But even there the margins are closing in and the trilling on a mountain top is now as likely to be a mobile phone as a sky lark. In the modern world remoteness is not about distance, but about information.

However, communication can distance as well as bring closer. We rely increasingly on constant access to information and other people. Why bother to try and keep track of a colleague's changing phone extension and address, instead simply keep a link to her WWW home page? How many people do you know of only through email contact -- do you even know their physical address? The corollary to this is that when the communication network is down everything is lost. Donne said "no man is an island". The issue we face is not so much whether we are connected to the maine, but that we are linked everywhere with a spider's web of interconnections. When we are isolated the threads of our lives come quickly apart.

When we unplug our laptop computer from the network and set off for the day on the train, we are stepping into remote territory just as much as Marco Polo or Dr Livingstone -- our colleagues may be only a few miles away, but we are cut off from them and from our familiar information world.

The contributors to this book are a bit like the cartographer Richard of Haldingham, who at the end of the 13th century made the Mappa Mundi which now hangs in Hereford Cathedral (UK), like the engineer Isambard Kingdom Brunel building iron bridges and ships which shocked his contemporaries, or like the 'navigators' who cut the canals which criss-cross the countryside around Birmingham and Huddersfield. They are all looking for ways in which people who are remote, in the modern sense, can understand the world they are in and work together despite distance.


top | intro. contents | notes | book home page

The Structure of This Book

Most CSCW (computer-supported cooperative work) research and commercial groupware systems assume that participants will have continuous access to relatively high bandwidth communications. This is acceptable for office based staff, but increasingly computer supported work is done out of the traditional office based environment: whilst travelling, at clients premises or at home. Such workers are often separated from their colleagues and their corporate information repositories, so the need to support the cooperative aspects of their work. Yet with their communication channels poor, intermittent or of low bandwidth -- how can they co-operate?

The chapters in this book approach this vital question from a variety of view points. On the one hand one can seek to improve the available communication by use, for example, of packet radio and compression techniques -- thus reducing the imbalance between the office based and mobile worker. On the other hand, others accept the available communications infrastructure and look for cooperative applications which fit within its limitations. The contributors are drawn from industry and academia and encompass both practical experience and theoretical analyses. The book is divided into five broad sections:

As with any such division there is significant overlap. In particular, most of the 'technology' chapters also address specific application areas. After considering each chapter in turn, we will return to the overall problem from a different perspective.

Firstly, chapters 1 and 2 address the broad social and economic context of remote work. Neal and Mitchell are writing from different sides of the Atlantic, but there is less difference in their messages than one might imagine. For example, in both can be found an emphasis on the more prosaic aspects of telework such as the use of basic, extant technology. The current barriers to teleworking are not a lack of global video-conferencing facilities, or even the limitations of the WWW. Instead, the primary problems are organisational attitudes and having too many settings to get right on the modem. Neal focuses on the concept of the 'virtual office' which may be in a jet at 20,000 feet above the Pacific Ocean, or in your own living room. It is quite interesting to contrast this with the concept of the 'virtual distributed home' in chapter 11. The effects of information technology seem to be a blurring of the edges of our previously compartmentalised lives. The icon of western post-war architecture is little boxes: office blocks, blocks of flats, suburban estates. In a post-modern world, IT seems to be shattering the substance of these boxes whilst still leaving the form intact. Mitchell primarily addresses the needs of mobile workers. His advice is very down-to-earth and is no doubt influence by his involvement with the UK Department of Trade and Industry teleworking survey. It is a message that will stroke chords with many who have to implement teleworking with non-IT expert staff.

The next three chapters deal with specific areas where technology has or can have an impact on cooperative activity. In chapter 3, Madigan looks at the construction industry where workers on site need to communicate with the design team, the client and remote experts. He describes solutions both at the site office and for the worker on site. In chapter 4, James and Savill describe the introduction of electronic information systems to Community Care Workers. They contrast the experience of the original paper-based systems with two generations of hand-held systems. The systems do not make use of sophisticated communications, but instead the information about the patients forms the focus for cooperation. Traditionally the sales force of a company have often been mobile and are often physically remote from the factory which manufactures the products they sell. However, whether or not they are physically remote, the sales force are organisationally remote -- lacking direct contact with the shop floor. In chapter 5, Ainger and Maher describe how modern cell-based factories are more cooperative than controlled and make it possible for the sales force to view directly the loadings of the factory and thus make more informed sales deals. Furthermore, information from sales can drive the production schedule -- a two-way interaction.

In chapters 6 and 7 the attention moves to the individuals involved in remote cooperative work. In a 'perfect' world we might like our cooperative systems to have infinite band-width communication channels and instantly see the effects of one another's actions. In the real world the technological solution is to improve hardware and algorithms to approach this image of perfect CSCW. Thimbleby and Pullinger focus on what each individual user sees of the system. If they cannot tell the difference between a 'perfect' system and the on they actually have, then the system is 'practically perfect'. They investigate some properties of a system which are necessary for this to be the case. Chapter 7 also looks at CSCW from an individual perspective. Many cooperative systems focus on sharing of information. Thomas argues that this will only work well if the system is designed from the 'inside out' -- effective personal management of information is the prerequisite of effective cooperative information systems.

The next three chapters look at various theoretical and practical aspects of the infrastructure of remote collaborative applications. All three are considering situations where there are no permanent wired network connections. In chapter 8, Busbach describes the design of a system, Task Manager, which facilitates the activity coordination of people working apart. Even with 'perfect' connectivity, the individual and organisational requirements for effective coordinated activity are far from trivial. In addition, the system must cope with periodic disconnection of individual's machines, yet restore a consistent and timely overall picture when connection is resumed. In our own contribution, chapter 9, we build on this theme of disconnection, considering the common requirements of mobile workers and teleworkers and especially their access to shared information. It focuses on two activities: retrieval and resynchronisation. Retrieval addresses the problem of what parts of the shared information are to be downloaded on to the user's local machine and resynchronisation considers the problems arising when users update the same information whilst working at separate sites. In chapter 10, Davies et al. consider the telecommunications and mobile computing needs of field workers in the electricity supply industry. These workers operate over large areas and need to communicate with each other and their central base. A failure in cooperation does not lead simply to poor or inefficient work, but can easily lead to loss of life. The chapter considers the relation between the typical work practices and the low level software, hardware and communications infrastructure required to support those practices.

Finally, the focus moves to the home. In chapter 11, Sloane takes a wide look at the changing understanding of the idea of the home. In previous ages the home was the focus for all types of work and leisure activity (and incidentally no strong distinction between paid and unpaid work). However, in industrial western culture the word 'home' has become somewhat debased, often no more than a stopping-off place between the activities of real-life. However, this is changing and the home is once again becoming the focus for the full range of human activity. Sloane calls this resurgence the new oikos. This is not simply a return to the past, a retreat from the uncertainty of post-modern society. Instead, he looks to the way in which information technology is opening up the boundaries of the home, making it a point of rich interconnection and interaction with the wider environment. In chapter 12, Eisenstadt et al. look at a particular manifestation of this. The Open University has a world-wide reputation for successful distance learning. However, one of the key features of OU modules has always been the Summer School, where students get a chance to meet (in the flesh) each other, their tutors and lecturers. Most OU students would vouch for the importance of this experience in motivating and adding to the breadth of their studies. However, a significant number of students have real difficulty in attending such residential sessions, even once a year. In this last chapter, we share some of the experiences of a Virtual Summer School, which attempts to recreate some of the atmosphere and dynamics of this occasion using electronic communications.


top | intro. contents | notes | book home page

Framing the Problem

The individual chapters address specific problem areas and technical issues. In order to get a broader perspective on the problem we use part of a framework for understanding CSCW introduced in a previous book in this series (Dix 1994a).


Fig. I.1. The CSCW framework


Cooperation requires two or more participants. These are labelled P in figure I.1. An important part of cooperation is direct communication and this is indicated by the arc between the participants. The study and facilitation of this communication is central to much of CSCW: email, video-conferences etc. However, this computer mediated communication is not all there is to CSCW.

In addition, the phrase cooperative work implies that the participants usually have some objects upon which they are working, whether physical or electronic. These artefacts of work are denoted A in the figure. Depending on the nature of the artefacts one or more of the participants will be able to control, modify or affect the artefacts. Furthermore the participants will normally be able to perceive the effects of their actions (feedback) and possibly those of others (feedthrough).

The ability to observe the effects of other people's actions is crucial to many cooperative situations. Imagine moving a piece of heavy furniture up some stairs. As well as talking to one another you will also feel one another's movements through the thing you are moving. This communication through the artefact may often be more important than the direct communication.

Now consider remote working. These channels of communication and control are torn apart (figure I.2). You can think of the work in the chapters in this book as in various ways bringing these two parts back together.


Fig. I.2. Broken links of cooperation


The use of information technology to bridge the direct communication gap is fairly mature and we see in various parts of the book references to the use of video-phones, mobile telephones, CU-SeeMe and email for asynchronous communication. Where these means are not already used (for example in the systems described by James and Savill), the difficulties are not in the availability of suitable technology, but in the choice of an appropriate mechanism (e.g. email or voice mail) and the human issues surrounding their adoption.

Possibly the greater proportion of the work reported in this book concerns the sharing of information. That is the lower part of the diagram. There are two sides to this problem. First, the availability of information to individuals. For example, the principle argument of Ainger and Maher is that improved access to factory floor information can make the sales force more responsive and competitive. One of the crucial issues here is ensuring that appropriate information is available locally. Hence the focus on retrieval and caching in our own chapter and on replication in Busbach and Davies et al.'s chapters.

Although this is clearly useful and important for the remote worker, is it really about cooperative work? The sharing of information comes because of feedthrough, when people are aware of and respond to the effects of one another's actions. In the sales situation the information from the factory floor must be timely, that is feedthrough of the factory staff's actions to the sales force. Equally important in order to schedule their work, the factory staff must become aware of the promises made by the sales force to the customers. Similarly, the 'electronic hard hat' is an attempt to bring the environment of the site worker into the virtual environment of those at other locations.

So where does that leave the community care workers described by James and Savill? They work largely alone during the days, with no electronic communication. The only interaction with their central site (and so indirectly with one another) being the nightly upload and download of patient records. The key lies in the pace (Dix 1992) of the cooperative tasks. In the case of electricity supply workers, the cooperative nature of the tasks has a rapid pace, in the order of minutes or seconds. The repair worker about to climb a pylon needs to know what whether or not the control room has shut off the electricity supply. A few hours before there may have been no fault, in another hour the electricity will be on again. The important thing is whether there is electricity in the wires now. For such workers continuous wireless communication could be a matter of life or death.

In contrast the cooperation over the treatment of patients in the community takes place over weeks, months and years. It is equally critical that information is timely and that different workers dealing with the same patient communicate and cooperate over her. This is evident from the cases cited in James and Savill's chapter. However, in the context of community care timely means correct within days not minutes and nightly connection is sufficient.

But what happens when the pace of required cooperation is greater than that allowed by the available communication channels? Often the only solution is to restructure the tasks so that less cooperation is required. However, the very apparent intractability of the problem makes it technically challenging. Clearly something has to 'give' in such cases. Feedthrough is limited to the pace of communication and effective feedback can only be achieved by replicating shared information in different places, with the attendant risk of conflicting simultaneous updates. Where technology cannot remove these problems we must accept them and design systems which help people to working despite their lack of contact. This is conundrum is central to Thimbleby and Pullinger's concept of observational consistency and our own analysis in chapter 9 of divergence and merging of shared data. It also appears as a theme in both Busbach and Davies et al.'s chapters.

Previous work on the CSCW framework has emphasised the importance of the link between direct communication and the artefacts of work, in particular deixis, the referencing of work objects in conversation, for example, speaking whilst pointing at things. In the area of mobile work these links are often particularly weak: direct communication may be by mobile telephone, or even video over ISDN lines, but the software which supports shared applications and data is completely separate. Indeed, during the Virtual Summer School the students were given mobile telephones as their normal telephones did not have the capacity for both data and voice transmissions. Even if advances in fundamental communications technologies remove such restrictions, the design of effective integrated environments for remote cooperation remains an open and complex problem.


top | intro. contents | notes | book home page

Future Directions and Further Information

We are in the middle of a period of rapid social and technological change in the areas addressed by this book.

Various occupations have always demanded a semi-nomadic lifestyle: sales, maintenance and executives. However, the efficiency of modern transport and telecommunications means that the distance over which it is possible to carry out such jobs has increased and so, following a sort of Sod's law, have done so. Organisations are increasingly trans-national, and even where they operate in smaller managerial units, the supply chains will usually be of national extent.

Not only are people mobile within their jobs, but the rapid redeployment of staff between offices leads to longer and longer commute distances. Coupled with this, in many countries multi-career households are the norm. Moving after a job change is no longer an option for many and week commuters are now common. One of the editors lives over 100 miles from his 'work place' (by no means the place where all work is carried out) and we know of married couples working on opposite seaboards of the United States.

At present the major barriers to even greater teleworking are principally managerial, organisational and social. The economic benefits are well accepted and technology is adequate (both editors can word-process, fax, and access email from home). However, these barriers are rapidly disappearing as management-by-objectives supplants management-by-presence (a lesson UK companies have found hard to learn). Furthermore, companies find they need to be lean to compete and the proportion of contract staff is steadily increasing.

The changes in technology over recent years have been dramatic. There will probably always be a price difference between desktop and mobile computers, but there are few applications which cannot be run on the road. Digital mobile telephony has matured and ISDN will soon be available in all major western cites. However, note this is not universal availability; in many countries and in the countryside everywhere provision at least of terrestrial telecommunications will lag. Information poverty is on the agenda. It is unlikely to be addressed by market forces alone and positive action may be necessary. For example, in Scotland the Highlands and Islands Initiative put pervasive ISDN into the most sparsely populated region of the UK, but was only possible with large European Union grants. It is unlikely that such socially-oriented funding will be available for all areas. Perhaps the only hope for such areas in the developed world will come from gentrification as affluent telecommuters take their jobs into the 'remote places' and so increase the demand for and hence supply of information services.

One of the theses behind the development of this book is that home-based working and mobile working share many characteristics; most significantly the relative paucity of electronic connectivity. While this is currently the case it may well not be a reasonable stance in five years time. The Internet revolution has transformed the information status of the typical home. It may soon be the case that the difference between the home and office is purely one of location and social climate.

The situation for truly mobile work is less clear. It is unlikely that wireless communications will be able to catch up with the rapidly increasing expectations for network bandwidth. Hence, as is true of the homeless of all ages, the nomad may be most information poor of all. There are efforts to address some of the problems of disconnected operation, for example, the development of Internet addressing to allow temporary connection away from one's base network. However, given the huge number of portable computers the facilities commonly available (rather than in the research laboratory) are pitiful. For example, object linking mechanisms behave very badly when files are copied between machines. The models built into such mechanisms presuppose that information permanently resides in the same (logical) location. This is problematic even in a fixed network and disasterous on the move.

Keeping up during a period of change is difficult. During the final stages of production of this book a special issue of the journal Distributed Systems Engineering (Blair et al. 1995) was dedicated to the issues of remote computing. There have also been two recent conferences on the technology and issues of mobile computing (IEEE 1994; USENIX 1995). For CSCW issues in general the best places to look are the alternating biennial ACM CSCW and ECSCW conference proceedings, although the number of CSCW books is also multiplying rapidly. Also for a broad introduction to CSCW and HCI in general see the editors' own text book on Human--Computer Interaction (Dix et al. 1993). There also now an increasing number of books specifically on teleworking, including a recently published collection of teleworking experiences collected by Andrew Bibby (1995). As befits the area, there are various web sites to visit. Links to some can be found from the book's own web page:

http://www.hcibook.com/alan/books/cscw96/

This also contains abstracts of each chapter and a complete copy of the bibliography in several electronic formats.

The issue of access to information and to one another is the essence of the chapters of this book -- similarly we hope that its very existence addresses this issue too: communicating ideas and concepts to foster a greater understanding of the problems and opportunites that exist.


top | intro. contents | notes | book home page

Notes

1. Images of the earth:
You can see live (half-hour update) meteorological images of the Euro-African hemisphere at:
http://www.ccc.nottingham.ac.uk/pub/sat-images/meteosat.html
LandSat and EarthSat both have an Internet presence. At the time of publication their pages contain only stills and project information, but no live feed (at least not public access!):
http://geo.arc.nasa.gov/esd/esdstaff/landsat/landsat.html
http://www.earthsat.com/
The US Geological Survey EROS Data Center keeps an online catalogue of data sets on the Earth's land surfaces:
http://sun1.cr.usgs.gov/glis.glis.html
All the above can be reached through the book's web page.


top | intro. contents | notes | book home page
maintained by Alan Dix