Improvement of automatic time series data validation and outlier detection in a real-time context
France, French Republic
SUEZ SMART SOLUTION offices located in Talence (Bordeaux suburb)
Data Science/Computing
Water, Utility, Environment

SUEZ puts digital technologies at the heart of its innovations to support all stakeholders involved in resource conservation: local authorities, companies, (industry, property managers) water, water treatment and waste processing services users. This is the core business of SUEZ SMART SOLUTION, a dedicated subsidiary that gathers 250 experts in fields such as environment, technical IT and data sciences. We collect and process our data and those of our clients and use them to enhance our expertise in the field of environment as a whole.
Over the last decade, SUEZ SMART SOLUTION has developed a leading edge expertise in real-time anticipation (RTA) of river and sewer systems, based on a dedicated platform called AQUADVANCED URBAN DRAINAGE. The RTA of such systems relies on the combined use of monitoring data, rainfall forecasts and numerical models to compute the hydraulic condition, and anticipate its behavior. This tool aims to rapidly provide accurate information in case of flooding events, before they occur, to help the authorities take the best decisions and minimize risks. The cornerstone of this system is data, gather by the application, from the field and from numerical simulations, which feeds the different modules. The quality of the results and the trust placed in them is directly linked to the quality of the input data.
There are multiple data sources, each with their own specificities, making it difficult to generalize data validation models. The purpose of this internship is therefore to make the different data processing algorithms in AQUADVANCED URBAN DRAINAGE more reliable, generic, and adapted to business challenges.
To do this, it will be necessary to rely on the methods present in the application, to integrate the new processes developed by our expertise and research center in Bordeaux, the Lyre, and to adapt them to the specificity of a real-time environment, while dealing with the constraints inherent to it.

Within that framework, the internship will have the following objectives:

  • Review existing methods, the ones developed by the LyRe, as well as the state of the art of existing methods in this field.
  • Test the different methods in a real-time environment on variables with different specificities.
  • Identify the strengths and weaknesses of each method according to the objective.
  • Apply these solutions on practical cases and evaluate the potential gain compared to current solutions.

If positive, prepare and implement this solution in our development platform.


The internship will have a total duration of 4 to 6 months and will be split in 3 successive parts:

  1. Preliminary phase: 1 to 1,5 months
  • State of the art on the existing methods.
  • Selection of a set of solution to be tested.
  1. Evaluation of the retain solution in term or performance and benefits: 2 to 2,5 months
  • Definition of test cases.
  • Run on the test cases and analysis of the performance and benefit.
  • Identify the strengths and weaknesses of each method given an objective.
  1. Integration in the real-time platform: 1 to 1,5 months
  • Development of a connector prototype and test.

Signal processing, time series analysis and machine learning.

Knowledge of hydrology or water quality processes would be a plus.

Programming languages, preferably Python.

Fluent in English or French.

Attracted by challenges, good interpersonal skills, good adaptability, proactive spirit.

  • Resume
  • Cover letter
  • Unofficial transcript
  • Recommendation letter
Yes (if validated by Embassy)

Company details

We’ve been experts in water and waste management for 150 years. We operate on 5 continents, on which SUEZ harnesses all its desire for innovation to achieve an efficient and sustainable management of resources throughout the world. Our company supports its customers as they change from a linear model, which over consumes resources, to a circular model, aiming to recycle and recover them for future use.