Anatomy of an Alarm & Monitoring System

The increasingly complex systems required on board today can easily lead to information overload for the crew operating them. Here, the president of Palladium Technologies, Michael Blake, explores the most effective ways to present alarm and monitoring data through the careful design of graphic user interfaces.

Yacht monitoring and alarm systems (AMS) must evolve to meet the demands of the ever-growing complexity of today’s yachts. In older systems, which only monitored a few sensor points, with less sophisticated systems, it was acceptable to have simple alarm notifications.

 

Let’s take a look at the anatomy of a properly designed AMS for the yachts that are now being built and start by breaking down the functionality of these systems into three distinct areas, which are:

 

  • Monitoring
  • Alarm processing
  • Reporting

 

The first category, monitoring, is the more visual aspect of these systems. The human brain has about 100 billion nerve cells or neurons, yet it has been proven that we can only hold seven digits or groups of information at one point of time in our working memory. The designers of monitoring systems must understand this limitation, as ‘information overload’ is very easy to achieve when too much data is presented to the user in one go.

 

By designing the graphic user interfaces so that only top-level information is displayed, and incorporating a simple methodology of drilling down deeper for more detail, the operator is helped to avoid the confusion of data overload. In other words, screen designs should eliminate the complex heavy use of data on any individual screen, while data should be grouped in clusters of no more than seven, correlating with our brain’s limitations of visual comprehension.

Data should be grouped in clusters of no more than seven, correlating with our brain’s limitations of visual comprehension.

The use of colour is also very important in the presentation of information as it has a powerful impact on how we react. As an example:

  • Red – increases blood pressure and stimulates the adrenal glands.
  • Yellow – stimulates the brain and can make you more alert and decisive.
  • Green – relaxes our muscles and helps us to breathe deeper and slower.
  • Blue – lowers blood pressure, has a soothing effect and stimulates the pituitary gland, which then regulates our sleep patterns.
  • Brown – brings feelings of stability and security.
  • Grey – a colour of evasion and noncommitment.

 

Designing these user screens to display useful information should be considered an art form in human interfacing. Much thought should be put into understanding the users and their environment, along with the data that is needed to do their job efficiently.
Alarm processing, the second function, is more sophisticated than just reporting high bilge water or temperatures. Today, these same AMSs observe thousands of sensor points, which require greater alarm filtering so that only real and valid alarms are presented to the operator. These systems must be semi-intelligent, emulating the way the engineers on board would think. For example, classification societies require alarming on transom doors while underway, but it is annoying to have this same door alarm while at dock or anchor. Building upon logic, they might want to say, “If the speed over ground is greater than two knots, and the transom door is open, then alarm”, which would then address this alarm requirement.

These systems must be semi-intelligent, emulating the way the engineers on board would think.

Providing a simple, easy-to-use scripting utility that allows the chief engineer to build simple or complex alarm logic, as shown in the example above, puts more power over the AMS back in the hands of the experts. It also eliminates the nuisance alarms, which ultimately causes real alarms to be overlooked.
Not all alarms are critical alarms, and it is best to have pre-alarms or cautionary alarms, which warn the crew of a possible developing issue. This also allows the crew to be proactive in solving developing issues before they become catastrophic. Having the ability to set multiple levels of alarms per sensor is essential in the management of the complex systems being monitored.

Relating to my statement of the brain’s limitation of seven elements of data in working memory, alarm points should be grouped by systems. For example, when an alarm occurs, if the first indication was the system, such as ‘bilge’, then the user can make an immediate decision if this alarm group requires immediate action. If so, then ‘drilling down’ will uncover the exact alarm(s) in that group that need action.

 

Reporting is the third aspect of an AMS that should not be neglected. Trending of historical data can uncover anomalies that might be missed in the details. We have experienced where plotting over time such areas as the coolant and exhaust temperatures for the port and stbd engines have revealed potential problems not obvious in the real-time information. This visual discernment is a strength that we all have, which allows us to isolate unique patterns from a picture, using inference.

 

Since there is a great wealth of knowledge to be gained in this historical data, we are firm believers that our systems should maintain at least five years of recorded sensor data, captured at least once per second, and 10 years of alarm history. A proper system should also provide operational reports, such as a snapshot of a group of operating sensors, once every four hours. This can help the chief engineer when maintaining his operating logs, and provide important historical back-ups.

 

In summary, the AMSs we use today are responsible for keeping these complex systems all functioning properly. In the past this was thought of as simply a safety issue, but we believe there is another corresponding function, which is to help predict potential failures before they occur.

We believe there is another corresponding function, which is to help predict potential failures before they occur.

Share your thoughts

No Comments

Leave a Comment: