PhD on characterization of abiotic stress of trees using AI methods on acoustic signals

Back to jobs listing

INRAE presentation

The French National Research Institute for Agriculture, Food, and Environment (INRAE) is a major player in research and innovation. It is a community of 12,000 people with 272 research, experimental research, and support units located in 18 regional centres throughout France. Internationally, INRAE is among the top research organisations in the agricultural and food sciences, plant and animal sciences, as well as in ecology and environmental science. It is the world’s leading research organisation specialising in agriculture, food and the environment. INRAE’s goal is to be a key player in the transitions necessary to address major global challenges. Faced with a growing world population, climate change, resource scarcity, and declining biodiversity, the Institute has a major role to play in building solutions and supporting the necessary acceleration of agricultural, food and environmental transitions.

Work environment, missions and activities

The MSCA project GreenField Data (IoRT Data management and analysis for Sustainable Agriculture) will mobilize fourteen doctoral candidates on various themes related to data for agroecology. This job offer is about the PhD G - Characterization of abiotic stress of trees using AI methods on acoustic signals.

The thesis will be carried out in the unit Integrative Physics and Physiology of Trees in Fluctuating Environments (PIAF) studies how living trees (in forests, cities, or agricultural areas) function, react, and adapt to major environmental factors, especially those impacted by climate change, such as drought, heat, cold, and wind (mechanical stress).  

Their goal is to understand the mechanisms of tree resistance and resilience at various scales (from cells to forest stands) to:  

- Identify resilient tree genotypes or ecotypes. 

- Propose sustainable management methods for tree systems and agroforestry.

Context: 

Abiotic stresses (e.g. frost, drought, wind) cause significant damage to natural and cultivated plants, which is expected to increase in the future with increasing climate variability (extreme climatic events). The detection of acoustic emissions is a promising way to measure continuously and non-invasively the damage affecting plants. Different sources of acoustic emissions have been identified (e.g. air bubble formation in conductive tissues, cell lysis, mechanical rupture, see references below) generating acoustic signals with their own characteristics. The analysis of the waveforms (amplitude, frequency, etc.) allows them to be discriminated under single stress conditions. However, to date, no study on a set of stresses (succession or interaction) has been carried out, and since plants are permanently subjected to different stresses, the use of this technique remains limited (in time, e.g. period of water stress, or in space, e.g. altitudinal limit). This case study therefore aims to better characterize the acoustic emissions generated by a single constraint and by their interactions, in order to ultimately develop a tool capable of measuring damage under natural conditions. 

Objectives: 

This case study will focus on two complementary parts: (i) analysis of acoustic signals to extract relevant information from it (signal quality), (ii) comparison of classified acoustic measurements with ecophysiological reference measurements in cultivated sites with different stress modalities (e.g. agroecological orchards and vineyards along natural gradients). The characterization of the acoustic signature will make it possible to measure the damage generated by different climatic hazards and to better understand the physiological mechanisms of resistance to abiotic constraints. The acoustic signature, integrated into the algorithm controlling the autonomous acoustic sensors, will make it possible to trigger alerts and an adapted response to these different climatic constraints. The design of a tool capable of measuring damage and, ideally, mitigating its consequences before it becomes irreversible is key to mitigate consequences of climatic stress. By providing a better understanding of the physiological mechanisms that plants develop to resist abiotic stress and, above all, their interactions, it fits to the challenge agroforestry and agro-ecology will face in the future. 

All the mentioned objectives can be listed as follow:

 1. investigate the potential of using acoustic emissions to detect and measure damage caused by abiotic stresses (drought, frost, etc.) in plants; 

2. develop a non-invasive method for continuous plant health monitoring based on reliable acoustic signatures; 

3. analyse the unique acoustic signatures of different abiotic stresses on plants by means of advanced analytics methods. 

This is a novel approach as previous research focused on single stresses, while in nature plants experience multiple or interacting stresses. These objectives will be achieved using the following work planning to grant their feasibility.

Work plan:

1. Conduct a literature review in data collection techniques used to collect the data for this project and ML techniques for multimodal datasets - month 1 – 6

2. Attend training on database of acoustic signals, applied stress and physiological indices collected in different woody species under drought and frost stress - month 3 – 6

3. Explore the diversity of signals and perform complementary experiments to finalize the training dataset - month 6 – 12

4. Develop a data analysis process based on machine learning for multimodal datasets and evaluate its performance and robustness of it results - month 12 – 24

5. Develop and implement an intelligent acoustic system and validate its results in the field (real-world data) -month 24 – 33

 

References :

Charrier G., Charra-Vaskou K., Legros B., Améglio T., Mayr S. (2014a) Changes in ultrasound velocity and attenuation indicate freezing of xylem sap. Agricultural and Forest Meteorology 185, 20-25.  Charrier G., Charra-Vaskou K., Kasuga J., Cochard H., Mayr S., Améglio T. (2014b) Freeze-thaw stress: Effects of temperature on hydraulic conductivity and ultrasonic activity in ten woody angiosperms. Plant Physiology 164, 992-998.  Charrier G., Pramsohler M., Charra‐Vaskou K., Saudreau M., Améglio T., Neuner G., Mayr S. (2015) Ultrasonic emissions during ice nucleation and propagation in plant xylem. New Phytologist 207, 570–578  Charrier G., Nolf M., Leitinger G., Charra-Vaskou K., Losso A., Tappeiner U., Améglio T., Mayr S. (2017) Monitoring of freezing dynamics in trees: a simple phase shift causes complexity. Plant Physiology 173, 2196-2207  Kasuga J., Charrier G., Uemura M., Améglio T. (2015) Characteristics of ultrasonic acoustic emissions from walnut branches during freeze–thaw-induced embolism formation. Journal of Experimental Botany 66, 1965-1975.  Lamacque, L., Sabin, F., Améglio, T., Herbette, S., & Charrier, G. (2022). Detection of acoustic events in lavender for measuring xylem vulnerability to embolism and cellular damage. Journal of Experimental Botany, 73(11), 3699-3710.  N. Chergui and M-T. Kechadi, “Data analytics for crop management: a big data view”, Journal of Big Data, 9(1), 2022, https://doi.org/10.1186/s40537-022-00668-2  Bansal, Y., Lillis, D. and Kechadi, M-T. A neural meta model for predicting winter wheat crop yield. Mach Learn 113, 3771–3788 (2024). https://doi.org/10.1007/s10994-023-06455-1  Ngo, Vuong and Kechadi, M-T., “Electronic Farming Records - A Framework for Normalising Agronomic Knowledge Discovery”, Journal of Computers and Electronics in Agriculture, 184(01), May 2021. http://doi.org/10.1016/j.compag.2021.106074

The thesis will encompass two periods. During one year and a half, the recruited candidate will work at University College Dublin, National University of Ireland, Dublin, Ireland (18 months). Then, until the end of the thesis, the candidate will work at INRAE, Clermont-Ferrand, France (18 months).

 

Due to the MSCA mobility rule, researchers must not have resided or carried out their main activity (work, studies, etc.) in Ireland for more than 12 months in the 36 months immediately before their date of recruitment

Training and skills

Recommended training: masters degree in computer science 

An applicant must have received the equivalent of 300 ECTS with a major in computer science, from which at least 60 ECTS corresponds to a master degree. The master degree must be granted by a university recognized by the International Association of Universities.

Required skills : advanced Machine Learning, data mining, and programming skills, interdisciplinary work, a taste for plant science and field monitoring will be appreciated, fluent (oral and written) english skills as the project operates in english language, team-mindedness, knowledge of the language of the host country may be considered a merit

INRAE's life quality

By joining our teams, you benefit from (depending on the type of contract and its duration):

- up to 30 days of annual leave + 15 days "Reduction of Working Time" (for a full time);
- parenting support: CESU childcare, leisure services;
- skills development systems: training, career advise;
- social support: advice and listening, social assistance and loans;
- holiday and leisure services: holiday vouchers, accommodation at preferential rates;
- sports and cultural activities;
- collective catering.

 

Salary: €3.100 in France complemented by a mobility allowance and a family allowance, adjusted by country

How to apply

Additional information: 

1266278343.pdf

List of PhDs - Greenfielddata (Ref G)

Application form: 

EUSurvey - Survey