Researcher
Research institution
Champion
Focus team
Topic
Project status
Year ended
2015
Project ID
201302
Why should I care about this project?

Is there a better way to display alarms on your HMI?  This project attempted, and failed, to create a method different than what was currently being used by members.  However, findings showed there are ways to improve alarm identification using methods already existing in most DCSs.

Abstract

Dr. Craig Harvey of LSU was attempting to develop a novel way to present alarm information. A series of interviews were conducted at different locations on how the operators utilized the alarm system. The general consensus was that operators were content with their current alarm displays, regardless of DCS vendor. Without a “problem” to solve, no novel way of presenting alarm information was developed. Project was terminated without achieving the intended results.

Objective

Initial studies have evaluated alarm rates and two different interface styles on operator performance. The fact that we have found differences in performance as a result of alternative interface designs suggests there may be other alternative interfaces for alerting controllers to alarms that ultimately could result in better overall performance. All current designs evaluated result in passive or reactive management; an operator in effect waits for an alarm and then addresses the event.

If we are to advance the HMI for alarm systems, the manner in which alarms and their impacts are presented to operators must be evaluated.

Driving questions

The question raised is whether different display types or even decision automation support may better facilitate active management of the system and prove more effective than the passive management used in traditional alarm monitoring.

Background

Previous research has focused primarily on alarms with specifics on alarm loads. Dr. Harvey and Dr. Ikuma have also briefly looked at simple display type, situation awareness, and subjective workload in the development of performance measures. This project expects to expand on that work by exploring how the operator is presented alarm information.

Deliverables

Only deliverables 1 and 2 were initially funded. Members wanted to preview the second deliverable prior to a decision on further funding. 

  1. A survey of industry HMI alarm displays, scenarios identified, and alarm handling needs analysis.
  2. New proposed HMI alarm displays.
  1. Defined interface displays and scenarios to be evaluated.
  2. Experimentation schedule and design.
  3. Analysis of performance metrics for each alarm interface display and a comparison of performance between display types.