Socio-technical systems engineering (LSCITS EngD 2012)

Socio-technical systems
engineering
Ian Sommerville

Socio-technical Systems Engineering, EngD course, May 2010 Slide 1

Rationale
• Socio-technical issues are now recognised by senior
managers in industry as a major issue in systems
design
• There is a pressing need to take socio-technical
issues into account when procuring, designing and
configuring complex systems
• However, methods of social analysis, such as
ethnography, have not been widely adopted in
industry
– Lack of expertise. Requires understanding of both the social
and the technical
– Cultural factors. Engineers should focus on technical issues

Problems
• Methods such as ethnography that rely on a situated
specialist examining an organisation are hard to scale
• Case studies of previous systems are retrospective
analysis techniques
• Little or no explicit support available for influencing
the design of a system based on a social analysis of
the work
• Inter-disciplinary incompatibilities
– Social science vs engineering perspectives


Socio-technical systems
engineering
• Embed socio-technical analysis into the engineering
processes of procurement, specification, design and
implementation of systems
• The aim is for socio-technical analysis to be carried
out by engineers and others involved in procurement
and design, rather than specialists
• Integrated with other systems engineering processes,
taking technical issues into account


STSE requirements
• Practical methods of social analysis that return useful
results in a reasonable time
• Guidance for non-specialists on how to use these
methods to study work
• Integration with processes and methods used in
systems engineering
• An evaluation framework that demonstrates the value
of socio-technical systems engineering


Practical social analysis
• Based on workplace studies
• Informal ethnography
• Action research / Co-realisation
• Guided ethnography
– Ethnographic viewpoints
– Patterns of cooperative interaction


Informal ethnography
• Observers simply go into the workplace and observe
work as it is actually practised
• Minimal previous training so low-cost of
implementation
• Reveals some of the most obvious issues of
cooperative work and work practice
• Problems of observer variation, inconsistent
coverage, etc.
• Suffers from same problems of scaleability
• BUT – better than nothing!

Action research
• A social science researcher is embedded in the
development team and applies methods of socio-
technical analysis as part of that team
• The goal of the researcher is to communicate socio-
technical issues to the team who then explicitly reflect
on these and how they should influence the design
• Problems of availability of specialists who can be
action researchers and inter-disciplinary
communications


Co-realisation
• A technical specialist is trained in methods of social
analysis and is embedded in a development team
• They apply social analysis methods and translate the
results directly into system design advice (and
implementations where appropriate)
• Avoids communication problems between social
scientists and engineers
• Problems of finding technical people sympathetic to
this approach and scaleability


Guided ethnography
• A framework is used to guide engineers and
managers in carrying out fieldwork
• More likely to lead to repeatable results and better
coverage than informal ethnography
• Based around:
– Ethnographic viewpoints
– The use of modelling notations to supplement field notes


Ethnography in requirements
engineering


Social viewpoints
• Ways of looking at a fieldwork site with a view to
understanding how the work is done
– Distributed coordination
• How do people coordinate their tasks as part of everyday work?
– Plans and procedures
• How are the objects in the workplace used and how is their use
governed by organisational policies and rules?
– Awareness of work
• How are activities organised to make work visible?


Social concerns
• Mechanisms to generate questions for each social
viewpoint
• Cross-cutting all viewpoints
– Paperwork and computer work
– Skill and the use of local knowledge
– Spatial and temporal organisation
– Organisational memory


Concerns and questions
• Skill and the use of local knowledge
– To what extent have standard procedures been adapted to
take local factors into account?
• Spatial and temporal organisation
– Does any data have a ‘use-by’ date?
• Paperwork and computer work
– How do forms and other artfacts act as embodiments of the
process
• Organisational memory
– How well do formal records match the reality of how work is
done?


Patterns of cooperative
interaction
• A means of communicating information about how
people interact with each other through and around
technology
• Developed as a resource that enabled the
generalisation and reuse of previous ethnographic
studies
• Represent patterns of work that are commonly
observed and their significance
• Provide a basis for fieldworkers to know what to look
for when observing the workplace
– Help address the problem of how to get started in
ethnography

Components of a pattern
• Essence of the pattern
• Why used?
• Where useful?
• Design implications


Examples of patterns
• Artefact as an audit trail
• Working with interruptions
• Collaboration in small groups
• Receptionist as a hub


Problems with ethnography
• Focused, primarily, on co-located work
• Excellent method for examining ‘work in the small’ i.e.
work as practised by a small team
• Not as effective for studying ‘work in the large’ i.e.
work across an enterprise or organisation
– Practical problems of studying many workgroups that are not
co-located
– Need to study management as well as use of systems
– Communications between parts of the organisation are
critical


Studies of organisations
• Challenge for socio-technical systems engineering
• Enterprise systems
– Goals of senior management may be different from goals of
individual work groups
– Diversity of comparable work across organisations
– Organisational complexity
– Rythyms of work and organisational timetables
– Legacy systems, processes and culture


Systems engineering and
change
• New systems are often introduced into an enterprise
as part of a process of organisational change
• However, there is often constrained and limited
communications between the systems engineering
team and the change team
• One view of STSE is as a means to bridge the gap
between the engineers developing the software to
support new processes and the change team
designing these processes


Systems engineering processes


Change processes


Bridging the process gap


Sensitisation
• Making sure that the stakeholders in the process
understand:
– Why human, social and organisational issues are important
– Why they are not someone else’s problem
– Why there may be good technical reasons that mean social
‘requirements’ have to be compromised. Engineering
constraints are significant
– Key issues from other stakeholders


Constructive engagement
• Working with stakeholders to implement new systems
• No prescriptive model of how engagement should be
practised or what socio-technical approach should be
used
– Engagement at all phases from procurement to operation
may be useful
– Make use of whatever resources and expertise are available
– Research issue is to develop lightweight methods of social
analysis that can be used as part of systems engineering
processes


Evaluation framework
• Evaluation of any changes to processes and methods
is difficult as ‘repeatable experiments’ are practically
impossible
– For example, debate over the utility and generality of agile
methods

• Utility of the approach may be demonstrated by fewer
problems in deployment and use
– But how can it be shown these are directly related to the use
of STSE


Key points
• Socio-technical systems engineering has the aim of
incorporating social and organisational analysis into
systems engineering processes
• Conventional ethnography is inappropriate and we
need approaches that are designed for use by
engineers
• Guided ethnography may be used for analysis of
cooperative work that is co-located
• Enterprise-scale analysis remains a research
problem


Socio-technical systems engineering (LSCITS EngD 2012)

More Related Content

Similar to Socio-technical systems engineering (LSCITS EngD 2012) (20)

More from Ian Sommerville (20)

Recently uploaded (20)

Socio-technical systems engineering (LSCITS EngD 2012)