Living lab at OiEau: innovation at the service of sanitation - the REISAR project

How can we reach networks that are difficult to access using external and internal GPS? How can we map the most complex networks with unprecedented accuracy? How can we contribute to the optimised maintenance of these infrastructures, which are essential for the safe transport of wastewater to a treatment plant? How can we support the safety of those working in the field? These are just some of the questions that the stakeholders in the REISAR project and the trainers at OiEau, who are specialists in sanitation, hope to answer by combining their complementary areas of expertise.

Developed and funded as part of the ‘France 2030’ programme, this robotics project includes the transfer of technological building blocks from research to industry. It will be carried out in two phases. The first phase, which will run until the end of 2026, will be carried out by researchers, robotics and geolocation specialists, and OiEau trainers to ensure that the robot is well suited to its end use and the professions involved. The second phase will be dedicated to the industrialisation of the robot and the potential markets it could serve.

"In this project, we are moving towards proof of concept, which means that our robotic system must be tested in an environment that is representative of the final system. That is why we called on OiEau and why we are coming to your Training Centre, to your teaching platforms, to carry out tests in clean sewers. The aim is to test what we have produced in the laboratory to date, rather than waiting until the end of the project.

We are addressing the issue of inspecting networks that are accessible or below the accessibility limit, to avoid sending humans into places where there are health risks, such as those associated with gas.

The robot we have created is sent inside. It must geolocate itself and scan the environment while being remotely operated in a simple manner and from a long distance by users, hence the aim of involving OiEau's expert trainers. We also need to make the link with professional requirements.

The robot will provide data, thanks to a complete 3D view, which will then be analysed in relation to the standards to be met in order to detect defects, for example.

We are not replacing human expertise at all in this area. The aim is to offer a kind of twin system and to protect personnel from dangerous locations. Indeed, it is difficult to recruit in these fields!

In addition, this robot will be very useful in meeting the requirement to map networks and ensure accuracy levels of between 40 and 30 cm. These are difficult to achieve outdoors with GPS, and indoors it is even more complex! This data will enable networks to be mapped, identifying the arrival of rainwater in sewerage networks and the resulting overload at treatment plants. This will prevent contaminated water from going directly into the environment or into the phreatic table. It will also enable repairs to be optimised.

The REISAR project has many interesting scientific aspects, such as geolocation, understanding environments and robotic control in sliding mode. The data will be made available to the scientific community fairly quickly. These research topics are related to the courses offered by our engineering school. The aim is also to feed them with this type of project. We then adapt the educational content based on what we are doing in terms of research. Hence the interest of being a teacher-researcher,‘"says Vincent VAUCHEY, Head of ’Transfer and Valorisation" within the Research and Innovation Department of the CESI engineering school, the main coordinating structure for the project.

Damien MENEUX, Robotics Manager at Pilgrim Technology, the partner in charge of creating the robot, its architecture and all the associated mechanics, explains how it works and the challenges it will help to address. "The REISAR robot, initially called Amphibot, is highly innovative because it uses a worm screw propulsion and therefore has no wheels. It is designed to travel in environments such as sand, mud, grass and water, in aqueous environments. It moves forward by constraint and therefore relies on its flooring to move forward, using the material. It uses its screws to move forward. The more liquid it is, the better, unlike conventional wheeled robots that get stuck. This robot floats, and once it is on the water, its screws act as turbines, like boat propellers. It uses the material it pushes. The more difficult the conditions are for ordinary rover robots, the more this one will thrive!

We know how to inspect storm water networks when there is a current. However, when there is a blocked section with stagnant water or sedimentation or sand, robots are unable to navigate it. This is what makes REISAR innovative and different from existing solutions.

The tests we carry out at OiEau are interesting because they are completely safe. This means we can focus on them and vary the water level, for example.

Another major advance is automation. Thanks to Pilgrim Technology's expertise, the technical complexity is handled behind the scenes to provide a very simple experience for operators. All you have to do is lower the robot into the pipe, press a button, and it takes care of the rest. An hour or two later, it comes back up with a full report and an impressive amount of data, ready to be used."

Charles-Edouard BOUCART, Technical Director of TRAAK, which specialises in geolocation systems in constrained environments, explains the technological approach adopted. "The main difficulty with indoor and underground geolocation is the absence of a GNSS signal. TRAAK's expertise lies in combining different existing technological building blocks, such as Ultra White Band, which enable short-range geolocation. In the REISAR project, we are involved in two aspects: retrieving precise positions with our trackers and transmitting the position using LoRa technology, which allows data to be communicated where it is needed.

This combination of technological solutions is quite sophisticated and still requires a more comprehensive configuration in order to obtain the robot's optimal position. Thanks to tests carried out at the OiEau site, we were able to experiment with a new method of penetrating materials underground. We captured data at depth and then transmitted it to the surface. This experiment enabled us to carry out a series of tests, the post-processing of which will help to improve our algorithms."

Data retrieval for an optimal user experience and communication between robots and operators is at the heart of discussions with Conscience Robotics, also a member of the consortium. This company, which specialises in robotic AI, is involved in LoRa communication and Human-Machine Interface (HMI) aspects.

Thomas Dupuis, Mechatronics Engineer representing the company, explains the approach. "For the HMI, we have an interface that is common to all the machines we have developed or integrated. We therefore have this unique technological building block. We are working on adding a few specific elements for the REISAR project, and addressing two key points.

The first is the user experience: being able to position the robot on a map, track it in real time and send it simple commands. During this exploration mission, we will be able to track its position in real time on the map, see where it is moving and, thanks to the data transmitted by LoRa, retrieve information on the size and shape of the pipes, for example on anomalies, with precise positions. These pipes will then be able to appear on a map in 3D. This data will be recorded in various formats for use by staff and sent to design offices to check whether or not repairs are needed.

The second is underground wireless communication. This is already well advanced, but there is still work to be done to enable optimal retrieval and better exchange of LoRa information. Improvements to the antennas will enable even better communication. We still experience losses at times. The aim is to limit these as much as possible. The LoRa communication part is really very innovative. It will be used in other projects later on, because it is a very useful building block that had not been of any use to us until now."

For all the players in the consortium, which is a shining example of 100% French collaboration, this first series of tests in real conditions has been very conclusive.

Each participant has been able to capitalise on a wealth of data and identify areas for improvement. The correction phase has therefore begun. Regular technical meetings allow them to exchange and pool their progress, with a view to converging on the same goal: optimal preparation for the second phase of testing, scheduled for early 2026, always at the OiEau Training Centre's educational facilities.

Interviews conducted on 10 September 2025