Interpreting data: a human-technology partnership
A large amount of data is now available to us all, both on a private and professional basis. Every day, we make decisions based on incomplete information, recognise patterns and act on other behaviours that a computer would and could not do. So, while computers are a good tool to help us find out what's going on in the world, using them in isolation is not the only way. This article looks at how to make sense of what computers are telling us and how to combine our human abilities and their technological strengths to develop effective, informed communications.
Data and information: what’s the difference?
Data is ‘raw’, non-interpreted information. Information can, conversely, be described as ‘interpreted data’. For data to become information, it must be meaningful to a human operator, in terms of both the task at hand and its overall context.
For example, the computer message, ‘error 468’ is data, while ‘error GPS 1 is off line’ is information that a human user can understand and act upon. To achieve meaningful communications, it could be argued that we need to deal with information, not just data. Unless we understand the meaning of the message, there is little point in having communication (at least where humans are involved).
Information management - how to avoid overload
New technologies bring with them increased communication possibilities and better access to data. Today, the challenge is to make relevant data/information available at the right time. If there is ‘too much’ data, or if it is presented in an inappropriate way, extra time and effort will have to be spent on interpreting the data’s basic content, rather than analysing its implications for the present situation and using it to inform future actions.
The right information therefore has to be available, but not in such large amounts that we become overloaded by it. If machines were smart enough to run without human interaction, they would know what information to present and when. Currently, humans remain better than machines at data interpretation, unless they drown in too much of it at once. You need to manage the amount of information available to you by focusing on the task at hand and knowing what you need to perform and when. It is all too easy to be sidetracked by too much information coming in all at once.
How much information do you need onboard?
Settings on bridge equipment should always be appropriate to the situation you are in. Lower ranges on the radar, for example, when closer to land or a large object, as well as choosing the band according to current weather conditions. Consider, too, who is on the bridge. Is it just you, or is there a whole team?
Different people may need different sets of information – the Pilot, Master, OOW etc. What might work for you may not be the ideal setting for others. Standardisation is often hailed as a solution, but it does not mean you will always have things your way.
The more notice you have before an action needs to be performed, the more information you can (potentially) have on your screen/s. Yet the closer you get (in distance or time), the more you need to unclutter your screen. You need to make sure you have both time to think and time to act. As people gain experience, they spend more time examining the situation at hand and less time considering multiple potential options. Novices, on the other hand, spend more time thinking about the options and less on comprehending the situation. As your experience grows, your “library” of situations and options will also expand, and you can recognise what to do quickly and more easily.
When you do not have enough time to think things through, you will have to make a trade-off, usually between thoroughness and efficiency.
To save time, you may well have to take a shortcut or find a work-around. Be aware of this and plan well in advance, while you still have the time.
THE RIGHT INFORMATION HAS TO BE AVAILABLE, BUT NOT IN SUCH LARGE AMOUNTS THAT WE BECOME OVERLOADED BY IT
Trustworthiness and accuracy
A lot of information today is presented in a graphically appealing way, which may mislead us at first. We humans judge trustworthiness like this when we meet people, and also when we “meet” new information. If it “seems” and looks trustworthy, we consider that it probably is. However, this may not always be true.
We must manage the available time and resources to try and find time to check the information out. Flaws in data and information may come from:
- Origin – i.e. they exist within the system from the beginning
- Installation or maintenance - for example, the systems are not correctly integrated
- Operational error
- External manipulation – this is more common than you might think.
There will always be a risk when basing a decision on incomplete information. AIS is one innovation that has given us many benefits, but also some new risks. If, for example, you make a decision to pass a ship based on the AIS data on the next port of call or the ship’s speed, that data may be faulty due to programming or incorrect incoming sensor data. It will therefore lead to what will look like a bad decision after the event. Always be wary, and double check your information any way you can.
Check, check and check again
There’s a reason you were told as a maritime student to always use more than one source of information. There’s a reason, too, that the COLREGS state you should check: “by sight and hearing, as well as by all available means appropriate in the prevailing circumstances and conditions…”.
The hardware may have changed, but the basic principle remains the same. Technology is not always trustworthy on its own, and human judgment may sometimes be based on incomplete data. Using both technology and human instinct in combination will help strengthen the approach and ensure a safer operation.
Author: Margareta Lützhöft
Deputy Director, National Centre for Ports and Shipping, Australia