The next idea she discusses is the well-meaning attempt to help manage the information overload problem is the idea of information filtering or trying to anticipate the information needed by the operator. The bottom line of this approach is that it puts the operator back in a reactive, rather than proactive operating stance and influences the performance of both the operator and the system. Finally, her comment is that user-centered design is not making decisions for the operator or doing things for the operator. So what is it? The answer to that question can be found in the book itself.
With all this new insight, what is the bottom line here? Today there are failures associated with breakdowns between technological systems and human partners. To fix this issue, the limitations, strengths and weaknesses of both partners must be understood. The strengths of both must be exploited. Having an operator waiting for an alarm before engaging in the system is a poor use of a human.
The human must work through a three-stage approach which includes awareness, perception of the elements in the environment, comprehension of the current situation and finally the ability to predict what will happen next -- projection of future status. To achieve this, operators need to be tracking trends and correcting them before the alarm state. When the trend or graphic has failed to make the operator aware, he or she needs to respond to alarms and correct the situation. The human need to comprehend is met by having displays that turn data into information.
For the operator to be performing like this, a control room environment is needed that is proactively supporting operator alertness, does not contribute to fatigue or distractions and supports the operators in all the tasks they have to perform. This includes non-DCS tasks such as report writing, MOC, performance improvement, training, etc.
With today's technologies, operators no longer have to look through a keyhole at infinite data. They can provide goal/task-based hierarchical information across large screen displays and an ergonomic desktop. This desktop will recognize the limitation of the human and will have no more than four working displays together with a permanent overview display. This means that the operator desk will also need to be ergonomic, and the DCS vendor's traditional console will be a thing of the past.
The HMI will move away from black backgrounds, colorful displays with small text, three- dimensional vessels and crammed information into a single schematic with hundreds more just like it. We have tried to differentiate what we have today with what we need by changing the name to High-Performance HMI. So performance must be monitored, measured and reviewed. Dependence on alarms to awaken the operator and attempt to reconnect him during abnormal events should be seen as an indicator of poor performance. The operator should always be in the loop, be predictive and prevent abnormal events by using trends and other more advanced graphic tools to put data into context. Alarms should be purely a backup feature.
Never again should an incident occur that has a major failure of the human due to overwhelmed operators who cannot respond to a simple level control alarm.
- REASON, James, Managing the Risks of Organizational Accidents, Ashgate, ISBN 1-84014-105-0.
- ENDSLEY, Mica R., BOLTE, Betty, and JONES, Debra G., Designing for Situation Awareness: An Approach to User- Design, Centered Taylor & Francis, ISBN 0-748-40967-X.
- The Explosion and Fires at the Texaco Refinery, Milford Haven, 24 July 1994, HSE Books, ISBN 0 7176 1413 1.
- The Esso Longford Gas Plant Accident report of the Royal Commission. Government Printer for the State of Victoria, No 61-Session 1998 -99.