What is Meduseo?
Meduseo is above all an information tool, an aid to save time and favor the right beach with complete peace of mind. Arriving on a beach and seeing the presence of jellyfish in the water, it must be admitted, it's annoying! Even if it is on its own, trust is often lost!
The ultimate goal of Meduseo is to provide a jellyfish weather service. In the same way that you will see the forecast of temperature, wind, or other, you can be informed of the presence of jellyfish on the beaches.
So predict the presence of jellyfish?
That's the goal, but of course before you predict, you have to be able to identify where the jellyfish are located every day. The simplest and certainly the most reliable approach you would expect is observation. Go to each beach, and assess the quantity of jellyfish. Some research groups have implemented this type of surveillance with high quality. Problem, only in France, more than 2400 beaches exist, so how to have as much information every day? One of the solutions is to have a collaborative approach where everyone reports what they see, but we all know how it goes, you have to motivate people, they will be all the more motivating during the holidays, but what about winter ? In short, even if the reliability is very good, not all cities will be represented and especially not every day.
Thinking about it, there had to be other ways to get this information. Inevitably, we first look at the existing models and simulations, but we must admit that for the moment this approach is unreliable at the country level. Jellyfish move mainly according to currents and winds, their presence on a beach can therefore depend on the orientation of the beach.
What if without knowing it, you were already sharing this information on jellyfish? The Internet is evolving, and the way people are using it too. Everything is shared, liked, retweeted, big companies (and others too!) Use powerful marketing tools to track you, know what you like, what you do. For once, couldn't these tools be used for more laudable purposes?
It is with this in mind that we have developed a series of tools dedicated to the extraction of relevant information on the basis of Internet 2.0 data in order to identify and report the presence of jellyfish on the beaches. In the end, you have nothing more to do, everyone speaks, shares information on social networks, and thus naturally feeds our model. Of course, in the state we can only make observations of the current situation and therefore we do not predict the presence of jellyfish the following days.
Finally it is good to specify that of course no personal information, private data is collected nor stored, we will not even start to do like the commercial ?!
What about reliability?
Ah! The big question! Difficult to have control markers for all coastal cities. But .... there are indicators! Sisi, remember, you must have seen in the newspapers "arrival of jellyfish on the beach of ...", "Beware of stinging jellyfish!", These are data that can serve as controls. So I allowed myself to make a dedicated page here, to give you some examples!
France, Italy, Spain, Australia, New Zealand
For all these countries we provide information on the presence of jellyfish. This list will grow, and will quickly include other regions where jellyfish are problematic like Japan or Africa.
You too can participate!
I told you, as it stands, the tool is sufficient by itself to work, but it will never be as reliable / precise as an observation on the beach. This is why you can also report the presence of jellyfish on the beaches via the application or the website. You will also inform other users, and strengthen the quality of information. You can also request the creation of a "certified" account (https://meduseo.com/info/46-certified), which will give you access to the site without advertising.
And what about the forecast?
We are working on it! I told you, before you predict, you have to observe, measure and assess the presence of jellyfish. We have now been able to collect reliable daily data for the past year. This forms the basis of the second phase of Meduseo, forecasting. We use the knowledge that we have been able to acquire in other scientific fields to, from information such as the weather, the forecasts, and the presence of jellyfish the previous days around, predict the appearance of jellyfish on the beaches at D+1. For this, artificial intelligence helps us a lot, whether it be approaches that have now become standard such as support vector machines (or SVM) or more complex approaches such as Deep Learning (fashionable at the moment!).
These forecasts are not yet available to the public and are still in the testing phase, but you will of course be informed as soon as the tool is available!
But where are there the most jellyfish ?
Who created this tool?
You have to go through this question! My name is Jérémy Deverdun, I am a research engineer in medical imaging, I develop tools, software and write a certain number of publications in the field of neurosciences. In my spare time I develop and improve Meduseo in order to make it available to as many people as possible (which sometimes explains my update times ...).
What you need to know (and this is often the case with me!) Is that I developed this tool some time ago now, in a very selfish way, for me! I found it unbearable to arrive on the beach and not be able to put your feet in the water because of jellyfish. I am therefore the first user! ... and also the first to complain when the site is not working! This therefore forces me to redouble my efforts to provide the most reliable information. If it is better, it should be better for you! :)
Unfortunately I do not share the source code that I developed, understand me, the tool is completely free, I take charge, I reserve at least that. On the other hand, I am perfectly willing to provide access via an API to part of the database within the framework of research protocols.
Thank you ...
To all those who will have taken the trouble to read this, and to those who by their help and support contribute to the existence and sharing of this site, and I also want to thank Midi Libre for helping to make Meduseo known through their article.