As we all know we have an enormous and continuous flow of information in our organization. People are talking, writing, taking photos, recording speech and taking videos. Information is produced continuously. Most of this information is just background noise but the relevant information is also in there somewhere. This information contains info of the current risk situation. People might have identified risks acknowledging it or they have taken a picture and don’t even recognize the risk yet.
A small fraction of relevant safety information is entered to safety information management systems today. This happens at least in the most advanced companies. But we are still missing probably 99,9 % of the relevant information that could be used to reduce the risk levels.
How to handle the information?
Problem that we first face in these semi controlled information flows is how to understand it all? How can we sort out the relevant information in real-time? These are really good questions. You can’t assume that a risk that is identified today can be fixed after one or two weeks without having a big probability for the risk to realize. Even though the person reporting the risk could protect themselves the others may not be aware of this risk.
We probably need to have some help to handle our massive information flows. Organizations have usually solved this with a very rational approach. The approach is to give people a list of predefined choices and this helps them to classify the observation. They can select from different lists and check boxes. Finally people feel that their observation doesn’t fit in and they just don’t book it at all. But the idea behind this is beautiful that we could use the reporting source to qualify the information for us and our content analysis wouldn’t need to be that sophisticated.
Using visual feeds
Another thing we all know is that one picture tells more than thousand words (or as we in Finland say also that it can confuse you more than thousand words). People are usually equipped with a camera and it’s the lazy way to report an interesting phenomena. This is the approach organization should encourage. A picture with perhaps some spoken added information or a video should be enough for reporting. It’s then up to the systems to interpret this information. It’s basically impossible for humans to effectively process the information your organization would produce if they would be given to opportunity to feed you with information as they are used to. But don’t worry. There is already some cutting edge research done to create an AI that can interpret pictures https://www.facebook.com/zuck/videos/10102456212502251/
There’s a huge potential to improve risk management process. As I wrote earlier, safety glasses will get smart and equipped with cameras and mics. The information collecting gets easier and more effective all the time. We need to focus on how to analyze that information as real-time and accurate as possible. This will save lives and it lifts the innovative forerunners to the NHL compared to the old school companies that will been seen as farm teams.