How can we observe animal species evolving in the wild without interfering with their lives? Observations are often delicate because they involve an intrusion likely to modify their behaviour or even their environment. A new innovative project, the S.I.E.L. project (Projet de Surveillance par Image des Espèces en Liberté), within the framework of the FSDIE (Fonds de solidarité et de développement des initiatives étudiantes) allows, thanks to the use of a new generation drone, to bring a solution to increase the duration of flight time, without disrupting the animal ecosystem: flights of more than 55 minutes and a 9-fold increase in video capture time.
Iy name is Robin Mathevet and is a 3rd year student in the Bachelor of Electrical and Automatic Energy (EEA) at Sorbonne University. Passionate about the design of drones, he is convinced that their use can contribute to innovative objectives. Because he has always been interested in biology, particularly through a course including a minor in life sciences, he wanted to put the technical advances of the UAV at the service of this discipline.
Robin Mathevet
He proposed his project to Olivier Adam, professor-researcher at the Institut d'Alembert (UPMC), working on the observation of cetaceans. Enthusiastic, he offered him a case of exceptional use: the observation of the behaviour of marine mammals in Madagascar. To test the prototype in a real situation, he accompanied him on his mission when the whales approached the Malagasy coast.
For four months, he worked on the design of the UAV, made the calculations and drew the plans. He anticipated the difficulties and needs in order to develop a UAV adapted to the conditions of the mission. He made it watertight and added floats so that it could ditch. He created a dual control system allowing the researchers to use the camera for their observations autonomously in parallel with the piloting.
With the researchers, they observed from the sky unprecedented interactions between whales and their young, without disturbing their behaviour or their environment.
At present, visual observations of animals are still favoured since, even if they have drawbacks, they are the only ones that make it possible to identify, in addition to their environment, the species, the individual(s), their activities and their possible interactions. They are fundamental if we want to understand them by listing and describing their activities as precisely as possible.
The arrival of aerial drones adds a new way of collecting data, whether for land, flying or marine animals. Whether they are winged or propeller-driven, we are in the very early stages of using these devices, particularly as tools for research programs. It should be remembered that drones are key tools for observation and surveillance. Their carrying capacity, stability, speed and low cost make them a flagship technology of our time. The only obstacle to their massive use is their short battery life. A drone flies for an average of 20 minutes while remaining motionless, which is not enough for many uses. Indeed, they need to reconsider the protocols, to evaluate their potential impacts and also to properly handle them. As far as winged UAVs in particular are concerned, they have certain drawbacks, in particular their low autonomy.
For research projects, UAVs are mainly used to describe specific geographical areas (with a high gradient, for example, or to have access to pedological information), or to observe animal species in their environment: this makes it possible to identify them, to inform about their habitat preferences, and possibly to monitor them.
Also, the S.I.E.L. project aims to provide a solution to significantly increase the duration of their flight time.
Objective of the project
The S.I.E.L. project aims to increase the flight time of multi-rotor propeller drones, within the framework of the cetacean approach code described in the charter resulting from an interministerial decree in Madagascar (Inter-ministerial decree n° 2083/2000). Multi-rotors used for civilian applications are currently extremely limited by their flight time; this disadvantage is amplified if sensors have to be transported and powered. On the other hand, at the aircraft's maximum flying time, one must consider the time needed to get to and from the scene of interest and finally, provide a safety margin. S.I.E.L. aims to find solutions to significantly increase this autonomy.
This project involves several students from several institutions. They have to design and build a prototype and then test it to evaluate performance. With different skills (mechanics, electronics, programming), they have to work together and respect the different stages to ensure the progressive progress of this project according to the chosen schedule. Therefore, animation is the key to this project by strengthening the links between them.
The design of the UAV takes into account the targeted application, through constraints on the choice of architecture: relatively small size for a boat take-off/landing, watertightness of the structure and engines for a departure/arrival at sea, ease of assembly/disassembly so that it can be done in an isolated place far from the surrounding towns.
The UAV has to fly at a great distance from the pilot to avoid interactions with animals, so visual piloting is impossible. To do this, a fixed camera is placed on the front of the UAV and transmits video in real time to a screen dedicated to the pilot to allow first person visual piloting (FPV).
The primary purpose of the UAV being to take pictures, it is equipped with a gimbal allowing to carry, orientate and stabilize on three axes a cam action type camera. The view of the camera is transmitted in real time to a co-pilot on the ground who can orientate it as he wishes with the help of a second radio control dedicated to the control of the gimbal.
Robin Mathevet has already designed numerous drones of all types, from the racer very light and fast to larger and more stable camera drones. In 2017, he worked on improving flight time by designing and assembling batteries that are more powerful than what is commercially available. His battery saved more than 60% of flight time - 23 minutes with a conventional LiPo battery versus 38 minutes with the prototype battery on an 1800g camera UAV, without impacting the performance or weight of the aircraft.
Tests in real-life situations
Professor Olivier Adam has found the S.I.E.L. project to be of great interest, as it deploys several complementary methods to study humpback whales, ranging from surface and underwater visual observations to acoustic recordings. Last year, he used a Phantom III propeller drone for the first time, but was disappointed by the very low autonomy (15 min), resulting in videos that were too short to be finally exploited. His solution envisaged for the 2017 mission was to use two drones sent one after the other, one after the other, to change the batteries of one while the other was in flight. This system is interesting but remains complex, insofar as it requires strong organisation on board the boat, and a high level of concentration on the part of the pilot, who has to manage two UAVs. The S.I.E.L. project offers the possibility of having a drone with an autonomy multiplied by 3, i.e. allowing the pilot to film the whales that interest him for about forty minutes.
Whales migrate to latitudes in Madagascar in July-August for their breeding season. They deploy several strategies to optimize their encounters. These interactions have already been widely described in scientific literature. But all observations have been made from boats. And underwater images are extremely rare, as it is very difficult to follow these cetaceans that are constantly moving and changing direction regularly. With drones, a new possibility of getting images has been born.
This is particularly interesting for studying humpback whales that live on the surface and sub-surface. Videos obtained by drone can be used to identify individuals and describe their respective behaviours (positions in relation to each other, aggressiveness, time of exclusion from the group). These data could be used in a much more concrete way in an ethological study, and allow to anticipate the victorious males who will have used more relevant strategies for the access to the female.
Researchers are also interested in the interactions of mother calves. In both cases, the duration of the observations must be as long as possible, as videos of less than 5 min (as was done last year with a Phantom III drone) do not allow for the description of scenes between these whales.
This test in real conditions has therefore enabled us to assess the use of this UAV by assessing the benefits of increasing the autonomy to more than 40 consecutive minutes while taking into account the sea conditions. The tests also highlighted the manoeuvrability during observations and the ease of deployment and recovery of the equipment.
All the interest of this project compared to previous UAV projects are the constraints that the UAV has to face. Being used in a context other than that of leisure, the device must be particularly reliable. In addition, the design will have to be particularly serious to avoid unpleasant surprises such as a battery that would not reach the 50 minutes hovering time objective or a poorly thought-out chassis that could resonate at the different engine speeds used during the missions it will have to carry out.
This project therefore has two major challenges: the creation of a UAV with exceptional flight autonomy and the creation of a machine for professional use.
Valuation of the project
The members of the project attach particular importance to its valorisation. S.I.E.L. is part of actions of valorisation, in particular in order to highlight the Sorbonne Universities and its training courses. Because this project was able to be born thanks to the FSDIE of Sorbonne University, the CFA of sciences, the departments of EEA and mechanical licences and the master sciences for the engineer. In 2017, it enabled Robin Mathevet to win, with two other students, Tracey Calme and Lucas Maigre, the FSDIE innovation prize.
Several actions are envisaged, such as, in particular, for the Physics and Engineering Sciences courses: organization of a 1-hour session (in the lunch hour for example) during which the inventors would come to present their prototype to the students and the results of this project. Because they are students in these courses, their approach aims at illustrating in a concrete way the concrete opportunities possible for these students, thanks to their knowledge and skills acquired at the Pierre and Marie Curie University. It is also a question of contributing to the valorisation of training courses in electronics and mechanics by testifying their experience, in particular to L3 students at the time of their choice towards Masters degrees.
The aim is for them to take advantage of this project to disseminate widely to other audiences. They are thinking in particular of the students of the MNHN, of Sorbonne Universities specialized in animal biology, ecology and environment, and also of the students of Paris Sorbonne on the social aspect with a project on Madagascar.
Beyond the purely scientific and technological aspect, it is also a project with a human dimension that they want to testify to by highlighting Madagascar, by bringing back exceptional images of this country and testimonies of the people they went to meet. They also want to give an account of the rich biodiversity of this country. In particular, they wish to contribute, at their level, to the conservation of cetaceans and the protection of their environment.
The general public must also be able to benefit from these advances: Every year, UPMC organizes stands to present its research activities during the Fête de la Science. It seems that this project could be presented there, either on a stand or during a conference for the general public.
Professor Adam is very involved in the valorisation of research activities: he participates in radio programmes, contributes to animal documentaries. In 2017, he organized a cycle of conferences on cetaceans at the Golden Gate Aquarium.
When Robin Mathevet is asked how he managed to produce unpublished images of whale sightings, he answers: "For four months, I worked on the design of the UAV, made the calculations and drew the plans. I anticipated the difficulties and needs in order to develop a UAV adapted to the conditions of the mission. I created a dual control system allowing the researchers to use the camera for their observations autonomously in parallel with the piloting. And so, on site, we were able to obtain flights lasting more than 55 minutes and multiply the video capture time by 9. Together with the researchers, we were able to observe from the air the unprecedented interactions between whales and their calves, without disturbing their behavior or their environment. »
How does he envisage the rest of his journey? "I'd like to start a startup that creates new generations of drones. I would also like to create a drone speed competition between universities and engineering schools within the Sorbonne University. In the meantime, I hope to continue making flying machines suitable for scientific research and improve the one already made during the S.I.L. project to enable it to take a sample of the whales' breath in flight. »
(Source: Sorbonne Universities)
To go further:
– international workshop dedicated to the use of UAVs for conservation," Unmanned aerial systems in marine science & conservation "
