Post doc – Using infrasound waves to supervise telecom infrastructures M/F

about the role

Your role will be to carry out a post-doc on ‘the use of infrasound waves for the supervision of telecoms infrastructures’, an activity linked to the FiberScope project selected as part of the France 2030 call for Deep Sea Projects.

It will enable Orange to identify the potential offered by Fiber Sensing technologies (fibre optic sensors) for detecting the environment around submarine cables, thereby improving preventive cable maintenance and optimising cable laying.

These sensors gather information about an infrastructure from the signals collected by the optical fibres running along it. In particular, each optical fibre can collect acoustic signals in the audible band or outside it, such as infrasound or ultrasound, along its entire length. 

In this study, we would like to continue exploring the possibilities offered by infrasound for inferring information about the environment and the use of an infrastructure, as demonstrated in the study [1].

You will be studying the feasibility of using this information to characterise events of significant interest in the management of our infrastructures. You will carry out this task using signal processing techniques in conjunction with Machine Learning.

You will also have to compare fibre sensing and dedicated infrasound sensors in order to characterise the differences in terms of sensitivity, directivity and range.

As part of the study, we are working with the support of a local company, a French expert in the manufacture of infrasound sensors [2], which is making several sensors available to us.

[1] Nicol R., Plapous C., Avenel L., Le Du T. (2023) Recording and Analyzing Infrasounds to Monitor Human Activities in Buildings, Forum Acusticum 2023, https://dael.euracoustics.org/confs/fa2023/data/articles/000641.pdf

[2] http://seismowave.com/home/products/infrasound-sensors/

The aim is therefore to take part in a campaign to collect infrasound signals using fibersensing and discrete sensors and to extract information that could be used to characterise events of interest. To do this, the aim will be to define a suitable methodology for capturing infrasound and optimise all the parameters to enable the expected characterisation and to implement it in the context of a demonstrator.

The study will focus in particular on the following questions:

• How can the extraction of potentially low-power infrasound signals of interest be improved using recent signal processing and/or machine learning methods?
• Is it possible and coherent to apply classical sound localisation methods to multi-sensor infrasound capture?
• Demonstrate the benefits of using discrete sensors and fibre simultaneously to determine events.

about you

 Skills (scientific and technical) and personal qualities required for the job

o A strong orientation towards innovation and applied research, combined with strong scientific curiosity and creativity.

o Good knowledge and basic algorithms in signal processing (and if possible audio and/or infrasound).

o Fluent English

o Good interpersonal skills and team spirit.

o Autonomy and thoroughness

 Education required (Master’s degree, engineering diploma, doctorate, scientific and technical field, etc.)

o PhD preferably in acoustics or physics

 Desired experience (internships, etc.)

o Experience in physical/acoustic measurements

o Experience in fibersensing

o Experience in computer development using Python (or MATLAB)

o Experience in Machine Learning

additional information

The subject proposes to explore an original approach with a high potential for innovation (scientific publications and patents).

In addition, the work will be carried out as part of the FiberSCOPE collaborative project.

Methodological approach :

1 State of the art on the issue

2 Evaluation of the digitisation solution with the reference sensors on loan

2.a Reproduce recordings in line with those from previous work using the new chain and new sensors

2.b Analyse the benefits of the three simultaneous sensors, in particular the possibility of positioning the sound sources.

2.c Adapt the recording protocol according to the previous conclusions

2.d Isolate new signals of interest

3 Participation in the recording campaigns

3.a Define a list of experiments based on the progress of the fibersensing recording campaign and the events planned.

3.b Adapt the previous recording protocol to the new conditions

3.c Create a multi-capture database

4 Combined analysis of infrasound and fibersensing signals

4.a Isolate the infrasound signals of interest, based on the results of the fibersensing signal studies.

4.b Assess the benefits of combined capture

5 Development of an algorithm capable of performing end-to-end extraction and analysis.

6 Evaluation of the algorithm’s performance

7 Production of a demonstrator

department

Orange Innovation brings together the research and innovation activities and expertise of the Group’s entities and countries. We work every day to ensure that Orange is recognized as an innovative operator by its customers and we create value for the Group and the Brand in each of our projects. With 740 researchers, thousands of marketers, developers, designers and data analysts, it is the expertise of our 6,000 employees that fuels this ambition every day.

Orange Innovation anticipates technological breakthroughs and supports the Group’s countries and entities in making the best technological choices to meet the needs of our consumer and business customers.

You will be part of a research team working on data capture, embedded and distributed Machine Learning and human/machine interaction. You will be part of an ecosystem that enables the concepts studied to be put into practice. The mission is to provide solutions for understanding the environment, in order to improve the quality of customer experience. For over a year, we’ve been focusing on Orange’s network and infrastructure operations.

contract

Post Doc

Read Full Description