Eiffel Team

Team

  • Research: ONERA The French Aerospace Lab (ONR)

  • Research: Université Pierre et Marie Curie (UPMC)

  • End User: Réseau de Transport d’éléctricité (RTE)

 

Autonomous damage Inspection in Power Substation

Eiffel Team is composed of two research laboratories ONERA and UPMC (ISIR) associated with the end-user RTE (Réseau de transport d’électricité). RTE is the company in charge of the French electricity transmission network, which operates more than 100 000 km of power lines and 2 700 substations. The target application considered is autonomous preliminary damage inspection in a power substation, which can highly benefit from the deployment of advanced Micro Air Vehicle (MAV) technology to reduce maintenance costs and increase availability and security of electricity network.

This application presents many challenging constraints for real-world MAV deployment, for which technological and algorithmical innovations are put forth. Power substations are highly cluttered environments where GPS localization is unreliable due to electric disturbances. Moreover, wireless communications can also be perturbed and therefore it is necessary for the MAV to be able to fulfill its mission autonomously.

Figure 1: Bird-eye view of the RTE substation "Le Chesnoy" (near Fontainebleau, France).

Vision-based localization and mapping will thus be the basis of MAV perception, with a focus on robustness to the environment characteristics: thin, linear and hollow structures (lattices and cables), reduced light conditions, unexpected or moving obstacles. Guidance and control algorithms will allow the on-line definition of exploration with real-time adaptation to the environment, including security guarantees regarding the avoidance of static or dynamic obstacles and tolerance to wind disturbances.

  1. The Benchmarking and free-style task will address the implementation of an advanced onboard avionic system to guarantee autonomous and safe navigation of the MAV in indoor static and dynamic environments. Based on the solutions developed during the qualifying stage, we will propose some improvements for the perception-control loop and we will demonstrate two new functionalities: autonomous exploration and dynamic avoidance of a non-cooperative moving object.

  2. Based on the autonomous exploration and mapping capabilities demonstrated in Task 1, we will introduce in the End-User Motivated Showcase task additional considerations for realistic constraints related to the power substation use-case. In particular, reliable visual-inertial localization and mapping with robustness to hollow and linear structures (lattices and cables) will be shown, as well as autonomous navigation under various illumination conditions (low light).

  3. The pilot experiment, which will take place at an actual RTE power substation, will combine the know-how demonstrated in Tasks 1 and 2. After reaching autonomously a waypoint while avoiding forbidden areas and unexpected obstacles, a local autonomous exploration and mapping of a delimited area will be performed before returning to starting point. The cluttered environment will contain many linear and hollow structures, and illumination conditions may be degraded. This final mission will demonstrate the interest for autonomous damage inspection and its impact on future crisis management.

Figure 2: Local autonomous exploration and mapping of a delimited area


Partner of

ICT Innovation for Manufacturing SMEs

www.i4ms.eu

News

30.01.2017

Best Drone-based Solution

2017-01-30 10:30 CET

On 2017-01-24 the EuRoC Team "GRVC-CATEC" was awarded the “Best Drone-based Solution” at the Eu parlament in Brussels. For more information, click...read more

Category: eurocNews

Displaying results 1 to 1 out of 19
<< First < Previous 1-1 2-2 3-3 4-4 5-5 6-6 7-7 Next > Last >>

Events

15.05.2017

EuRoC Showcasing Workshop

Date: 2017-05-15 till 2017-05-17

Type: Evaluation Meeting (partialy closed to the public)

Venue: Fraunhofer IPA, Stuttgart, Germany read more

Category: eurocEvents, eurocFutureEvents

September 2017
Mo
Tu
We
Th
Fr
Sa
Su
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30