Computer vision is a discipline that has made it possible to control different production processes in industry and other sectors for many years. Actions as common as the shopping process in a supermarket require vision techniques such as scanning barcodes.
Until a few years ago, many problems could not be solved in a simple way with classical vision techniques. Identifying people or objects located at different positions in images or classifying certain types of inhomogeneous industrial defects were highly complex tasks that often did not provide accurate results.
Advances in Artificial Intelligence (AI) have also accompanied the field of vision. While Alan Turing established the Turing test in 1950, where a person and a machine were placed behind a wall, and another person asked questions trying to discover who was the person and who was the machine, in computer vision through AI, systems capable of reproducing the behaviour of humans are sought.
One of the fields of AI is neural networks. Used for decades, it was not unitl 2012 that they began to play an important role in the field of vision. AlexNet1 , designed by Alex Krizhevsky, was one of the first networks to implement the 8-layer convolution filter design. Years earlier, a worldwide championship had been established where the strongest algorithms tried to correctly classify images from ImageNet2 , a database with 14 million images representing 1,000 different categories. While the best of the classical algorithms, using SIFT and Fisher vectors, achieved 50.9% accuracy in classifying ImageNet images, AlexNet brought the accuracy to 63.3%. This result was a milestone and represented the beginning of the exploration of Deep Learning (DL). Since 2012, the study of deep neural networks has deepened greatly, creating models with more than 200 layers of depth and taking ImageNet´ s classification accuracy to over 90% with the CoAtNet3 model. which integrates convolution layers with attention layers in an intelligent, deep wise way.
Turning to the relationship of modern computer vision models to AI, Dodge et. al (2017)4 found that modern neural networks classifying ImageNet images made fewer errors than humans themselves, showing that computer systems are capable of doing tasks better and much faster than people.
Among the most common problem solved by computer vision using AI are: image classification, object detection and segmentation, skeleton recognition (both human and object), one shot learning, re-identification, etc. Many of the problems are solved in two dimensions as well as in 3D.
Various vision problems solved by AI: Segmentation, classification, object detection
Classification simply tells us what an image corresponds to. So for example, a system could tell whether an image has a cat or a dog in it. Object detection allows us to identify several objects in an image and delimit the rectangle in which they have been found. For example, we could detect several dogs and cats. Segmentation allows us to identify the boundaries of the object, not just a rectangle. There are techniques that allow us to segment without knowing what is being segmented, and techniques that allow us to segment knowing the type of object we are segmenting, for example a cat.
Skeletal recognition allows a multitude of applications, ranging from security issues to the recognition of activities and their subsequent reproduction in a robot. In addition, there are techniques to obtain key points from images, such as points on a person´ s face, or techniques to obtain three-dimensional orientation from 2D images.
Industry segmentation using MaskRCNN5
One Shot Learning allows a model to classify images from a single known sample of the class. This technique, typically implemented with Siamese neural networks, avoids the need to obtain thousands of images of each class to train a model. In the same way, re-identification systems are able to re-identify a person or object from a single image.
The high computational cost of DL models led early on to the search for computational alternatives to CPUs, the main processors in computers. GPUs, or graphics processing units, which were originally developed to perform parallel computations for smoothly generating images for graphics applications or video games, proved to be perfectly suited to parallelising the training of neural networks. In neural network training there are two main stages, forward and back-propagation. During the forward process, images enter the network and pass through successive layers that apply different filters in order to extract salient features and reduce dimensionality. Finally, one or more layers are responsible for the actual classification, detection or segmentation. In backward propagation, the different parameters and weights used by the network are updated, in a process that goes from the output, comparing the obtained and expected output, to the input. The forward process can be parallelised by creating batches of images. Depending on the memory size of the GPUs, copies of the model are created that process all images in a batch in parallel. The larger the batch size we can process, the faster the training will be. This same mechanism is used during the inference process, a process that also allows parallelisation to be used. In recent years, some cloud providers have started to use Tensor Processing Units (TPUs), with certain advantages over GPUs. However, the cost of using these services is often high when performing massive processing.
Skeleton acquisition, activity recognition and reproduction on a Pepper robot6
CARTIF has significant deep neural network training systems, which allows us to solve problems of high computational complexity in a relatively short time. In addition, we have refined several training algorithms using the latest neural networks7 . We have also refined One Shot Learning systems using Siamese networks8. We also use state-of-the-art models in tasks such as object and human recognition, segmentation and detection, image classification, including industrial defects, and human-robot interaction systems using advanced vision algorithms.
1Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 25.
2Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., … & Fei-Fei, L. (2015). Imagenet large scale visual recognition challenge. International journal of computer vision, 115(3), 211-252.
3Dai, Z., Liu, H., Le, Q., & Tan, M. (2021). Coatnet: Marrying convolution and attention for all data sizes. Advances in Neural Information Processing Systems, 34.
4 Dodge, S., & Karam, L. (2017, July). A study and comparison of human and deep learning recognition performance under visual distortions. In 2017 26th international conference on computer communication and networks (ICCCN) (pp. 1-7). IEEE.
5He, K., Gkioxari, G., Dollár, P., & Girshick, R. (2017). Mask r-cnn. In Proceedings of the IEEE international conference on computer vision (pp. 2961-2969).
6Domingo, J. D., Gómez-García-Bermejo, J., & Zalama, E. (2021). Visual recognition of gymnastic exercise sequences. Application to supervision and robot learning by demonstration. Robotics and Autonomous Systems, 143, 103830.
7Domingo, J. D., Aparicio, R. M., & Rodrigo, L. M. G. (2022). Cross Validation Voting for Improving CNN Classification in Grocery Products. IEEE Access.
8Duque Domingo, J., Medina Aparicio, R., & González Rodrigo, L. M. (2021). Improvement of One-Shot-Learning by Integrating a Convolutional Neural Network and an Image Descriptor into a Siamese Neural Network. Applied Sciences, 11(17), 7839.
I´ m on my way to work and I hear an ad for a soft drink on the radio; during my break I see that my favourite singer encourages me to try it on social netwroks, he tells me it´ s great!; in the afternoon I go to the supermarket and I find a promotion of that soft drink which I can try it for free and there is also a 3X2 promotion; at night I´ m watching a series with my family and I see how the main character drinks that same soft drink with the brand clearly visible and shows an incredible satisfaction after drinking it… where is the limit between advertising and influence?
I am an adult with critical sense who can make the decision to consume a product or not, but… what about a child? Can we consider that children are free to make healthy choices taking into account all the advertising environment that surrounds us?
In Spain, 40.6% of children between 6 and 9 years of age suffer from overweight and obesity1, alarming figures similar to those of other countries such as United States or Mexico. The prevalence of childhood obesity in Spain is among the highest in Europe according to the WHO.
Today´ s lifestyle has changed drastically in recent decades and is believed to be responsible for the increase in overweight and obesity in all age groups and especially in childhood: children now consume more fast food and sugary drinks, eat away from home more often and spend less time eating as a family than previous generations. In addition, processed foods are more accessible than ever and are available in larger portions. Moreover, television and Internet use have led to a more inactive and sedentary lifestyle, as well as greater exposure to the marketing of products high in fat, sugar and/or salt (known as HFSS).
It is clear that to reverse this high prevalence of overweight and obesity in children, there is no single solution but it must be a set of actions aimed at reducing sedentary lifestyles and increasing energy expenditure in addition to improving consumption decisions towards healthier products, but, I ask again the question from before, can we expect a child to make healthy consumption decisions when in their daily life they have so many impacts of unhealthy products specifically aimed at children? According to a study by the OCU (Consumers and Users Organization), nine out of ten food advertisements aimed at children are for products with an unhealthy nutritional profile2 : cookies, breakfast cereals, industrial pastries, chocolates, enregy drinks. And many of them are advertised by influential characters or cartoons, accompanied by promotional gifts or collectible stickers that encourage repeat purchases and capture the interest of children, or endorsed by certain health associations.
In terms of advertising, there is some consensus3 that until the age of five, children are incapable of perceiving the differences betweent programming and advertisements, or that they do not begin to identify a persuasive interest in advertising until they are about eight years old. Not even after the age of eight it is guaranteed that minors will be able to identify messages as biased, since, as adults know, they tend to emphasize the positive aspects and ignore the negative aspects of the product.
In Spain, the PAOS Code was signed in 2005 with the aim of establishing a set of rules to regulate advertising and promotional activities aimed at children and to guide companies to comply with it. However, reality shows that children continue to be the target of many unhealthy food advertising and the figures for overweight and obesity continue to be alarming.
For this reason, the Ministry of Consumer Affairs intends to approve a Royal Decree regulating the broadcasting of unhealthy food and beverage advertising when it is aimed at children and adolescents up to 16 years of age.
The regulation that will start to be applied in this 2022 year will affect five categories of products that will not be allowed to advertise to children under 16 regardless of the nutrient content: chocolate and sugar confectionery products, energy bars and sweet toppings and desserts; pastry and biscuit products; juices; energy drinks and ice creams. For the rest of the product categories, a limit of nutrient content per 100 grams is established. In this case, they may be advertised as long as total and saturated fats, total and added sugar and salt levels remain below the limits established for each product. These limits correspond to the nutritional profiles established by the World Health Organization.
Advertising on television, radio, cinema and internet, social networks, websites or mobile apps will be regulated and there will be limitations on advertising in print media. There will be reinforced protection schedules in generalist television channels set from Monday to Friday, between 08:00 and 09:00 in the morning and from 17:00 to 20:00 hours in the afternoon, and on Saturdays and Sundays, between 09:00 and 12:00 hours, while the prohibition in children´ s television channels will be permanent.
Source: Consume Ministry
The intention of Royal Decree is in line with the recommendations of the European Commission in its Action Plan against Childhood Obesity and which is already applied in countries such as Norway, Portugal or the United Kingdom. In 2017, the European Commission published a report on the exposure of children to HFSS food advertising and marketing4. Some of the conclusions of this study were:
64% of food and beverage advertisements for children under 18 were for HFSS products.
A child under the age of 12 may be exposed to a total of 732 HFSS ads in a month.
80% of online HFSS ads are advertised on YouTube and 20% on traditional web pages
The most promoted category is sweet snacks.
Children see approximately 10 times more HFSS ads than health food ads in Romania, 6 times more in Sweden and 3.5 times more in Lithuania and Italy.
At the food industry level, there are also initiatives to adapt and improve this situation. This is the case of the EU Pledge initiative that promotes among its members the commitment, by January 1, 2022, in relation to the restrictions on the marketing of HFSS products, either not to advertise any food and beverage products aimed at children under 13 years of age, or only to advertise products that meet the EU Pledge nutritional criteria. The EU Pledge is currently adhered to by 23 companies that account for 80% of advertising expenditure in the EU.
The need for a regulation that regulates the advertising and promotion of unhealthy foods through all media that reach the child population is a reality. It is not only a matter of placing limits on food choices that can lead to health damage, but also of limiting the influence, incitement or suggestion of products in an unfair way, hiding their harmful condition, especially when the consumer cannot reasonably identify it.
The food industry also has a fundamental role in this task, both at the level of advertising regulation, as well as in the reformulation of existing products and in the research of other healthy and attractive options for children. From CARTIF, we continuosly collaborate with the food industry for this purpose,as in the projects PROBIOMIC (Design of new cereal products with probiotics adapted to optimal child nutrition through omic technologies) or TOLERA (Development of more effective and safer ingredients and foods, aimed at people with food allergies and intolerances), among others.
Caves were our first home but, have we stopped to think about how our ancestors felt in the cold mornings of winter? And in hot summer days? We may be surprised…
Humanity had had multiple and different homes. From the tipis of the american indians to the skyscrapers that flood nowadays the city of New York. Currently, buildings represents 40% of the energy consume and 36% of the greenhouse effects. Much of them, moreover, are from the 70s. Definetly, we need a change if we want to mitigate the climate change.
In the Palaeolithic, the first dwellings, in the form of huts made of animal skins and logs, protected our ancestors from the cold and wind. During the Neolithic period, the construction of villages with adobe houses provided our ancient inhabitants with habitable conditions. And all this without consuming a single kilowatt hour and using the resources that nature offered them to obtain certain conditions of comfort.
If we look at the evolution of buildings throughout history, we can see that adobe houses gave way to the dwellings of ancient Egypt, which were made of straw and wood. Ancient Rome introduced concrete and stone, as well as technologies such as the round arch, the arcade, the vault and the doem. Leaping forward to the Renaissance, this era marked an architectural breakthrough, including materials such as marble, stucco and tiles. Until the evolution towards the brick that makes up the majority of existing buildings. But despite the evolution in the use of materials… are we really improving our comfort conditions and the energy efficiency of buildings?
The answer today is that we need more efficient and smarter buildings, but what is stopping us froom changing the way we use buildings? Platón, in his myth of the cave, tells us that it is a lack of knowledge that hides reality from us. Extrapolated to the present day, the lack of useful and valuable information limits us when it comes to making more objective decisions, based on knowledge and reducing subjectivity.
To answer the question of how we improve the knowledge of buildings, the concept of intelligent buildings comes into play. According to the European Commission, an intelligent building is one that is connected, is able to interact with the systems around it, including users, and can be managed remotely. In other words, it has to behave interactively both with the building´ s energy sytems and with other buildings and even the users themselves. Furthermore, it changes its behaviour from reactive to pro-active to make efficient and effective use of its own resources.
The main enablers of smart buildings are new technologies. Firstly, the IoT (Internet of Things) which, in a nutshell, is defined as the connectivity through the Internet of common elements such as household appliances, cars, mobile phones, etc. It is this technology that makes it possible to turn a traditional building into a connected building, capable of providing data thanks to IoT sensors. Secondly, Artificial Intelligence, which uses data to extract knowledge; the same knwoeldge that, following Platon´ s myth, will guide us out of the cave. Artificial Intelligence is a technique capable of learning from data, extracting patterns of behaviour and predicting future situations. In this way, it is able to anticipate events and enable the building to act proactively. In other words, it is bringing human reasoning to buildings, but making decisions based on objective information.
At CARTIF, we have been working for years in the line of research for the transformation of current buildings into samrter, more comfortable and environmentally friendly buildings. Projects such as BRESAER are a clear example of this transformation. In this project, a decision-making system based on Artificial Intelligence has been developed. This solution allows the building to determine one hour in advance the energy needs to meet the comfrot conditions and to choose the available sources to heat or cool the building.
All this without forgetting that buildings are for us and, therefore, users must be the protagonists. Consumers must be better informed about the behaviour of the building, just as the building must adapt to the preferences of the inhabitant. For example, smart thermostats that learn our habits to ensure a comfortable temperature without the need to configure it. Or even detecting when we leave to switch off and stop consuming gas or electricity, which makes even more sense with today´ s prices. The example of this technology is part if the COMFOStat project.
In conclusion, smart buildings represent the perfect solution that combines today´ s better living conditions with the reduced gas emissions of old. Data and Artificial Intelligence generate the necessary knowledge that will have guided us out of the cave. If you still can´ t find your way, our door is always open to help you.
There is only one good: knowledge. There is only one evil: ignorance.
Since the beginning of the times they exist few things that we know they are going to happen with totally security. The world is full of happenings and alleatory events very difficult to predict, even for our partner artificial intelligence, which often seems to be a kind of panacea that can solve everything.
However, it exists something that I guarantee you its going to happen, and in a certain way, it is happening at this moment. We are getting old. Inside our life cycle, time goes by as we live through experiences, complete milestones and achieve goals. This time, even if we don´ t want it, little by little leads to a decline in both physical and physcological abilities, to the extent that perfoming daily tasks within the household is a real challenge, and sometimes even dangerous. At this moment doubt arises, am I looking for someone to help me or am I looking for a residential centre to live in?
For much people, their house is their independence symbol. Their house is that place where they can impose their own rules and were they don´ t have to be accountable to anyone. According to a study realized in 2020, in Spain exists around 4,849,900 people living alone, and inside this data, more than 2 million have 65 years or more. This is the 43.6% of the total. However, the odd thing of this statistic isn´ t only the high number of elderly that lives alone, but that this percentage has increase in a 6.1% compared to previous year. Therefore, it´ s assume that the tendence of society is living alone once you reach 65 years.
However, as I started to explain at the beggining of this post, it arrives a moment in which remembering some easy tasks, as it can be taking medicines at the right hour, can be difficult or even frustrating, and failure to remember can lead to a dangerous situation. With all these, it would be interesting counting with a person or system that reminds us taking medicines at time if we haven´ t, or remind us that we have to eat if we have forgotten, but without “annoying” us during the rest of day. This could involve a topic that is very much in vogue nowadays, home automation.
We could say that we start talking about home automation in the 70s, with several building automation pilot tests, but it was not until the 80s when it started the development of a commercial level for its distribution in urban households. At present, without going too much into standards and technological aspects, the following breakdown can be made within such a large branch as home automation:
Sensorization and data collection (If this occurs): It´ s about the first stage to have into account inside our home automation system. What we want to do is to collect data and events inside our household. We want to know if the street door has open for knowing if we have been robbed, our house temperature in case the heating needs to be turned on or presence in a certain room so that light is switched on automatically. All these can be reached thanks to technology, that monitorize the status of our household through a sensors network that measures physic parameters, as temperature, humidity or luminosity.
Actuators and implementers of action (then do this): Once we know what has occured inside the house, would come into play the second stage, we indicate to a socket switching on an electrodomestic, for example, or to a little engine to open up a door or window.
With all these, it is understood that people carried out their daily tasks following more or less established patterns. For example, a person entering his bathroom, closes the door and thereafter humidity inside the room starts increasing over standard levels, it can be deduced that it is taking a shower. Another example could be that is lunchtime and the temperature in the area were glass ceramic hob is located starts to increase, at the same time as the fridge and the drawers containing the species are opened. It can then be deduced that the person is cooking. Therefore, it is possible to track the tasks performed by an elderly person living alone using a home automation network that collects the events occuring in the house and an artificial intelligence (such as a neural network) that processes this data. Once the data acquisition stage has been completed, it would be interesting to integrate this information with the different telecare systems in the region. In this way, depending on the daily activities that are detected (or, alternatively, undetected), the telecare system can provide suggestions to the person or, if a dangerous situation such as a fall is detected, intervene in person.
From the Health and Wellnes area of CARTIF we seek to offer solutions so that older people can live as fully independently as possible for as long as possible. For this reason, one of our research lines it is focused on the contain of this post so that older people could stay at their homes in a totally operative and safe way. The theme that has been treated about home automation will serve to provide support facing the decrease of both physical and sensorial abilities. However, we are also working on solutions to improve the autonomy in households facing the physical deterioration through the development of technological assistants for the use of toilet and intelligent walkers.
To sum up, I want to emphasise that is very important to take care of the wellnes of our elderly and provide solutions that allow them to be fully active and to enjoy a healthy mind. Wether we like it or not, time is passing for all of us.
People hear a lot about the decline of bees, about the lack of pollinators, but what are pollinators? more importantly, what do they do for us and what do we do for them?
The group of pollinators is very wide, not only honey bees, which belong to one specific family. In fact, in Spain we have more than 1,000 species of bees from six different families, 75% of bees are solitary and live on the ground and with more than 20,000 species of bees in the world they have evolved and are organised in many different ways.
Other pollinators are birds, mammals and reptiles because plants are very clever – they have been on earth for many millions of years longer than we have! And that results in greater evolution and adaptation. Plants have developed sophisticated methods of attraction to achieve their reproductive purposes because the pollen has to reach its destination! They use everything from lizards to flies and bats, not forgetting air and water, which also help in pollination.
But if there are so many means for pollination, why are pollinators, in particular bees, so important? Well, at least, in this case, size and shape matter! There are all kinds of insects, small, fat, long, with very long tongues, in short, many sizes! Just as there are many shapes and sizes of flowers and pollen grains because plants are very sybaritic.
Plants have evolved in such a way that each one has developed its own system, some of them very exclusive, to avoid pollen from other plants, that is why there are so many smells, to make a first selection of “be my guests!”; and many sizes, some bees have to stick out their proboscis or “tongue” up to 20mm to get to the food; some, like the passionflower, have very large stamen and pistils, and can only be pollinated by large bumblebees; others are complicated, like the snapdragon flower, where the bee has to get inside as if it were a cave; and others are smaller, likedaisies, which need a small insect and are therefore pollinated, for example, by small black and yellow striped flies, which are the hoverflies. The sunflower is a large daisy and therefore needs a larger pollinator such as the honey bee. There are thousands of families of bees of different sizes, ranging from 4-5 mm to 30-35 mm (honey bees measure between 15 and 20 mm).
Both flowers, sunflower and daisy, have large yellow or white “petals” around them (which are actually ligule) and a bunch of little yellow flowers in the centre from which the seeds come out (which in sunflowers we call pipas). Next time you pass by a park, pick up a daisy and look closely at the yellow part, they are all little flowers! With their stamens, stigmas and all the parts of a flower that we hardly remember from when we studied them at school!
Cut of solitary bee nest. Source: Luis Óscar Aguado
Colours are another attraction mechanism for pollinators to detect them from far away! With our eyesight, all colours look the same to us, but with their special vision, they see the colours of the flowers differently. In the end, plants have made insects and pollinators evolve as pollen carriers, and in return, they provide delicacies in the form of fruit, seeds, pollen, nectar, etc. As you can see, there are numerous mechanisms to attract the right pollinator, so if the population of one disappears in a short period of time, the plant cannot adapt and even less reproduce.
Bee laying an egg on a bamboo cane. Source: Maria González
Many pollinators obtain food rewards from plants, but they do not feed exclusively on them like reptiles or birds, but bees do, they depend exclusively on plants for food. Both their larvae and the adult insect feed on floral products such as nectar and pollen. And as we have seen, not all flowers feed all pollinators.
So how can we help pollinators?
We can build them a shelter, depending on the family of bees they have different forms of housing or social grouping. Some build galleries in the ground where they live, so placing a pot in the window is enough; others in holes in logs or pieces of wood, even shells!. So a simple and space-saving way is to put up a hotel for solitary bees. It is very simple, consisting of a bundle of cut bamboo canes or a piece of wood with holes of different diameters between 5-25 mm, with a base, that is, that does not go through the wood and without cracks, as these are possible entrances for parasites and predators. There the bees will lay their eggs, which they will leave with the nectar or pollen collected and cover it up, and after a year the new bees will leave and go and look for another hole. So there is no danger of being stung. Bees don’t live there, they just lay their eggs, they are solitary, they don’t form hives, they don’t form communities, they sleep on flowers and branches.
Solitary bees don’t sting! Well yes, but not like honey bees, I mean, many bees die when they sting you, solitary bees do not form communities, as they do not have to defend it, they are not aggressive, they do not attack, they do not sting, this makes it a good idea to install this type of shelters in schools, children can get close and see how many holes are plugged, and count the bees they have helped.
Also, these bees only collect food for their own food and larvae, so they don’t need to collect so much, others need food to raise all their offspring, and others need food for the whole colony and humans, so if you disturb them with your hand, most of them leave, they don’t want trouble, attacking you costs them their lives, but others are more insistent because they need a lot of food and the closer you are to the hive, like the honey bees, the more aggressive they can become and attack.
Bee hotel with two holes fulfill located in CARTIF. Photo: María González.
This is why there are special regulations for keeping beehives, which are considered a type of livestock farming (beekeeping or honey farming) and have to comply with requirements regarding distances to populations, etc. It even has to respect the distance between them, since, if there is little food, it attacks other pollinators, such as solitary bees, as they are competing for food.
It is also important to have flowers all year round, for them and for us, there are many native plants with different flowering periods so that they have food all year round, in different sizes for bees of all sizes.
How do they help us? The best-known part of the work of pollinators is that of pollination. 80% of plants depend on them, and many of them are responsible for providing us with food. 75% of crops currently need pollination, and this includes the crops needed to feed livestock, so the production of livestock by-products also depends on them.
Pollinators module, module with flowers during all the year located in CARTIF. Source: María González
Pollinators are therefore our allies, and it is thanks to them that a large part of natural and urban plant communities and our food is maintained. Climate change, pesticides, agricultural intensification, the continuous clearing and mowing of parks and fields, including the proliferation of beehives, are causing their population to decline rapidly. Moreover, we still have a lot to learn about them, so it is very necessary to help them consciously and rationally, and not only at the level of experts, but also at the social level, all of us can help their conservation, taking care of pollinators and cultivating more flowers around us, since without flowers there are no pollinators and vice versa.
Bibliography: Curro Molina & Ignasi Bartomeus. 2019. Guía de campo de las abejas de España. Esditorial Tundra, Castellón. 250pp. 19’5 x 12’5 cm. ISBN 978-84-16702-77-0.
Although sometimes we forget it, forests provides huge benefits to the planet in general and to the human being in particular. They help us to mitigate climate change effects acting as carbon sinks and eliminating huge quantities of carbon dioxide of the atmosphere. The forests nourish the ground and serve as a natural barrier against ground erosion, ground movements, floods, avalanches and strong winds. Forests host more than three quarters of global terrestrialbiodiversity, and represents a source of food, medicines and fuel for more than one thousand million people.
But forests are seriously threatened by deforestation, climate change and fires. The advance of the agricultural frontier and the unsustainable logging causes 13 million hectares of forest to be lost every year. Climate change is allowing that plants and invasive insects species have advantages over the native species increasing their negative effects. It also exists a direct relationship between fires, deforestation and pandemics: the destruction of forests, specially the tropical ones such as the Amazonia, Indonesia or the Congo, makes possible that human beings get in touch with wildlife populations carriers of pathogens.
With regard to forest fires it has been noted that fires are becoming less frequent, but more destructive. Some of them, the most terrible, are the called “sixth generation fires“, and are ravaging the forests of the planet. This type of fires can´ t be fight and also they have the capacity to modificate the metheorology of the place where the fire is located. Against this type of fires it only works a defensive strategy, trying to direct it to non-populated areas and hope that the rain will help to control it. Not even areas that have hardly had any fires are not spared from this tragedy: 5.5 millions of hectares have burned in the Artic Circle in recent years. The Artic is warming twice as fast as the rest of the planet and, as a result, high intensity fires are starting.
It is clear that is fundamental to prevent fires and for that reason it is necessary to consider strategies that allows reducing forests vulnerability. Having a look at our nearer context, the European Unionforest strategy promotes the forest sustainable and respectful management with climate and biodiversity, intensifying the surveillance of forests and giving a more specific support to silvicultures. Becomes evident that is needed a better forestry management with emphasis in the protection and sustainable regeneration. However, we have a steady decline in forest mass as the “reforestation” process cannot compete with the deforestation rate in Europe. Furthermore, in Europe, data shows a large increase in forestry exploitation in recent years, which reducing the continent´ s CO2 absorption capacity and possibly indicating wider problems with the EU´ s attempts to fight agains climate crisis. Another paradox regarding forests within the EU is that a large part of them are privately owned by timber companies. As a result, the regular logging of these forests, coupled with the private nature of their ownership, makes public awareness and greenning even more difficult to achieve. Biomass loss from 2016 to 2018, in compared to the period from 2011-2015, has increased by 69%, according to the satellite data.
Spain, as it occurs to all the countries of the mediterranean area, is specially vulnerable to fires, given the scenario of drought and desertification, accelerated by the climate change. In Spain we have a large experience putting out forest fires: we collaborate in a international level and we achieve the extinction of 65% of fires in their outbreak phase (less than 1 hectare), although this sometimes produces the effect called “the extinction paradox” (which means that we lose the opportunity for small fires to clear undergrowth and thus encourage large and dangerous accumulations of fuel. In Spain 1,000 million euros per year are destiny to fire extinction, however, only 300 millions euros to their prevention.
The extinction is necessary and positive but isn´ t enough, it is necessary to invest in other measures (prevention, detection and recovering) that allows facing forest fires from a more wide and complete perspective. In this sense is very important to take advantage of new tools that offers recent technologies and scientific advances.
For example, the use of images obtained with drones and satellites and sensor grids joint with artificial intelligence techniques allows to detect fires faster and more accurately and is already underway several research projects in various countries: Bulgary, Greece, Portugal, Lebanon, Korea and much others. Even there are challenges planned for the European Spacial Agency for using satellite images and artificial intelligence in the detection of fires and other similar challenges of the NASA, H20.ai and Cellnex. Another interesting initiative is ALERTWildfire, a consortium of several northamerican universities that provides cameras and tools against fires to discover, locate and monitor forest fires. There are also commercial systems to detect forest fires, such as this one of Chile, that use Artificial Intelligence and several types of sensors or this one of Portugal.
Already in Spain, the Ecology Transition and Agriculture ministries have developed Arbaria project able to “predict” with a considerable hit rate where fires will break out.
Looking for a global approach in the prevention and management of fires the european project DRYADS have been launched, in which participates CARTIF. This project has as an objective the development of a fire management holistic platform based in the optimization and reuse of last generation socio-technologic resources. These techniques will be applied in the three main phases of forest fires:
In the prevention phase, DRYADS proposes the use of a real-time risk assessment tool that can receive multiple ranking inputs and work with a new risk factor indicator driven by a neuronal network. To create a community model adapted to fire, in parallel to the previous activity, DRYADS will use construction materials activated by alcali that integrates post-fires wood ashes for buildings and infrastructures resistant to fire. DRYADS will also use a variety of technological solutions, such as the Copernicus european satellite infrastructure and swarms of drone for a precise forest supervision.
In the detection phase, DRYADS proposes several technology tools that can be adapted to much of the needs of the project: use of virtual reality for the training, portable devices for the emergency services protection team, vehicles without driver -UAV (drones), UAG and aircrafts- to improve the capacity of temporary and spacial analysis, as well as to increase the coverage of the inspected area.
Finally, DRYADS will construct a new forestry restoration initiative based in modern techniques, such as agrosilviculture, drones for spreading seeds, IoT sensors that can adapt the seeding process in function of the ground needs and at the same time with the help of the AI to determine the risk factors after the fire.
The results of DRYADS project will be demonstrated and validated in real conditions in several forestry spaces of Spain, Norway, Italy, Rumany, Austria, Germany, Greece and Taiwan.
To sum up and as a conclusion, to fight against the forestry fires we have not only to focus in their extinction but also in a good sustainable management of the forest based in the prevention and introduction of modern techniques is essential to reinforce their resilience, the utilisation of the resources and their recovery capacity. This will lead to new opportunities for the rural environment, the biodiversity conservation and the fight against climate change. Let us hope that for once a time trees let us see the forest and we could avoid their destruction.
