home | resilience library | organizational resilience | IT and resilience - 1

IT and resilience - 1

(Image: Control panel from the Apollo Guidance Computer which flew Apollo 11 to the moon and back. It operated with the equivalent of 4 kilobytes of memory, a million times smaller than the 4 gigabytes found in many modern (2013) computers, cameras, memory sticks and mobile phones.)

There can be very few organisations that don’t have some form of IT – most have significant networks, a computer on every desk, individuals provided with laptops and tablets, and meeting rooms equipped with digital projectors for presentations. Software for word processing, desktop publishing, spreadsheets, presentations and communications is everywhere. Most companies have websites, many make use of the same technology internally via intranets.

We’ve replaced the typewriter, flipchart and envelope with digital equivalents, but is there more that our computers could do? Most of us have noticed that the packages we use have features that we’ve never explored, and we wonder where all our disk storage capacity disappears to. A comparison with the computer on board the Apollo 11 spaceship, (which guided the craft to the moon while having about a million times less computing power than our modern machines) makes it clear that we are using only a minute fraction of the computing power available to us.

In the 1980’s I worked in the field of industrial automation, and spent much of my time in control rooms. Later in the 1990’s I started to work more with senior managers, and I was surprised to discover that their use of technology was lagging far behind. Process operators had access to technology which was far more effective than their bosses many layers up in the organisation.

It is worth listing some of the facilities that these control-room systems provide, as I will be arguing that managers at all levels need something similar.

Process control system features

  • The system gathers information in real time from sensors and intelligent instrumentation on the production line.
  • It performs any data conversion required to integrate the information into a single database.
  • It will normally perform reliability checks on the information collected, highlighting any invalid signals.
  • Information is displayed in easily understood formats, such as animated mimic diagrams and graphs.
  • There are often features allowing an operator to zoom in on any desired area of the process to obtain more detailed information.
  • Information is updated in real time.
  • The system displays and logs alarms, alerts and warning messages, usually with an audible warning ensuring that the operator attends to the signal
  • A list of live alarms that have not yet been resolved can be displayed.
  • Historical records are stored from which any required report may be produced.
  • The operator may interact with the system by initiating sequences, adjusting settings, editing recipes, performing shutdowns, and other actions appropriate to the process.
  • Interlocks prevent operators from performing incorrect actions.
  • Facilities are provided for maintenance and diagnostics of the control system and the process plant.
  • Communication links are provided to systems at higher, lower and adjacent levels.

There are exceptions, but in my experience many managements do not have access to this quality of information, much as they would love to have it. Why not? I believe there may be a number of reasons:

1)      Managers may not be aware of the possibilities and benefits of applying process control technology to their own teams.

2)      Process control systems tend to deal with fairly static architectures – it is too costly and disruptive to keep changing the configuration of a production plant, and this tends to be done only after careful consideration and planning. By contrast, many managers have to deal with fast-moving situations, where information gathering has to catch up after decisions have been made.

3)      There may be a perception that data is harder to get at, as it is often buried in spreadsheets which have been created ad-hoc to perform a specific job, without necessarily considering the bigger picture.

4)      Management information is not emotionally neutral in the way that much (but not all) process data is. Reporting an error or defect may have an adverse effect on an individual’s career or bonus, and sharing information may be seen as giving away power. Some of the information that management require may be of a sensitive nature, particularly when it relates to individuals or the intellectual property of the organisation.

I believe that none of these reasons are insurmountable. I’m already starting to deal with the first one in this article, although nothing I write can be as good as some hands-on experience in an actual control room. The second and third points can be addressed by investment and planning at an early stage. An organisation needs more than a physical network – it needs an information infrastructure to run on that network that is capable of delivering similar features to a process control system across the organisation. Once the infrastructure is in place there will need to be a transition period during which information is ported across – this will be neither painless or cheap.

The fourth issue can partly be dealt with by using anonymised or aggregated information, but it also relates to organisational culture. If there is a culture of blame, individuals will tend to withhold or massage information regardless of the form it is in. Likewise trying to drive an organisation through arbitrary targets will create distortions in employees’ behaviour and the information they report.

Image:  The idea of an operations room for management was one championed by Stafford Beer in the 1970’s, and he succeeded in implementing one in Chile in 1973, despite a severe shortage of computing equipment.

Cybersyn Control room