Willing to contribute to real-world projects and transform mathematical challenges into value for the industry?
Find out the challenging projects you could work on:
The mathematical model behind the European day-ahead market for electricity is complex, with many non-convexities and combinatorial aspects. Different types of products are used in different countries, and two network models currently coexist in Europe.
N-SIDE has developed a tailor made two-step algorithm with a branch-and-cut approach to solve this problem, under demanding operational constraints.
This algorithm, called Euphemia, has been successfully running for more than 1000 days and is continuously improved.
As an intern, you will help us to face new challenges raised by the constant growth of the market and by the future extensions. There are many interesting topics to cover, related to the flow-based network model, the search for a unique demand price in Italy (called “PUN”), numerical difficulties, the gap estimation, etc.
In the context of production planning for pharmaceutical products, compare MIP formulation with constraint programming. MIP formulation already exists.
As an intern, you will develop an alternative constraint programming approach to the same business problem and challenge the two approaches in terms of performances (time, optimality) and sensitivity to small deviations in the input parameters.
Scala / Oscar environment is involved. Basic knowledge about mathematical programming, constraint programming and Scala is required, knowledge about Oscar environment is a plus.
- Machine learning
A large amount of data is the key for optimisation and decision-making.
As an intern, we need you and your data-mining knowledge to forecast one of these:
- Supply demand concerning care services
- Late on planned rides
- Energy market prices
The technology used is your choice as well.
From the input data sources to the forecasted data analysis, your part is really the computation core of that chain. Join N-SIDE to solve challenging machine-learning problems !
Edge is the framework we are building to support all our products. Built with Scala and Scala.js, it offers a powerful server and a modern and user-friendly web-based UI. While it is already used in production, there are still many aspects we would like to improve or tackle and that is where you come in.
- Advanced data visualisation
As an intern, your role will be to build advanced data visualisations (generic, reactive, interactive and editable) with state-of-the-art technologies (D3.js, Scala, Sass…). For example: network, time series, flow, map…
- Reactive collection system
As an intern, your role will be to design and implement a reactive collection system in Scala:
- similar to Map, Seq and Set (in both properties and ease of use)
- sorted or not, filterable, mappable…
- with content that can evolve with time (and corresponding events)
- unit tested and performance benchmarked
We are already using an early version in the edge UI that is meant to be replaced by your implementation.
- Supply Chain (Plants)
Strategic network design consists in optimizing the full distribution chain, from the production at the plants through the warehouses to the final clients. Typical recommendations of a network design tool are in/di-vestments in the plants/warehouses based on bottlenecks, production-storage-flow volume allocation, marginal profitability of the demands, etc. Given the integrated nature of the optimization, clients must be grouped strategically.
The grouping has to take multiple factors into account:
- the selling price,
- the network cost to reach the clients,
- the sensitivity of the client,
This grouping is essential for the quality of the optimization recommendations, as badly grouped clients would give meaningless results.
As an intern, your role will be consist in creating a client clustering algorithm for the purpose of strategic network design optimization.
- Smart Management of EVs fleet and Charging Stations
One of the main source of difference between Electric Vehicles (EV) and standard fuel-based vehicles comes obviously from the type of energy used. Compared to fuel, electricity
- can be produced locally,
- is unpractical to store on a high scale,
- has a lot of technical constraints for its transmission and distribution,
- contracts evolve from fixed price to variable pricing (hourly or quarter-hourly) in order to account for the increased volatility in the market price of electricity due to the increased share of renewables in the energy mix
These particularities lead to interesting topics for smart algorithms to help the key actors take the right decisions.
For fleet managers:
- Smart charging: depending on the production of renewable energy sources and the price of electricity, when should each EV be charged in order to minimise my charging costs ?
Optimal investment in local charging stations: How many charging stations should be bought in order to minimise investment costs while reducing risk of unavailability ?
Optimal investment in local electricity production and storage: How many photovoltaic panels and batteries should I buy in order to be able to supply my EV fleet on renewables ?
For charging station managers:
Dynamic pricing: how should I set the price of the electricity sold at my charging stations so that my revenue is maximised ? This will depend on the price of electricity on the market, the amount of energy coming from local production, the price elasticity, …
Pricing of flexibility: how to adapt my pricing so that drivers that are more flexible (willing to wait more in order to be charged for example) are incentivised ? Indeed flexible drivers can be charged at a lower price since charging can be delayed to when renewables are producing thus providing cheap electricity.
Optimal investment in local electricity production and storage: what investments should be done in order to reduce the price of electricity at the charging stations ?
And many more …
As an intern, your goal will be to design and implement a proof-of-concept addressing one or several of these topics. You will have the opportunity to apply your skills to real business challenges on a cutting-edge topic.
- Smart Routing of a fleet of Electrical Vans/Trucks
Today, the number of electric vehicles is increasing. The number of industries using such kinds of vehicles also increases. Some companies use electric vehicles to deliver goods or perform service tours. While the planning of such tours is a well-known problem, the electric vehicles change an important characteristic of the problem: the battery life and charging time cannot be ignored. Indeed, with fuel powered vehicles, the tank reloading time is insignificant with respect to the tour time. With charging times of hours, this assumption is not valid for electric vehicles.
- App for Online Flexibility Audit
Nowadays with the growing share of intermittent Renewable Energy in the mix, it is more and more important for large electricity consumers to be flexible and to adapt their consumptions on electricity availability.
However, it often difficult to assess what level of flexibility can be extracted from the different industrial processes of a plant, what are the constraints that limit the possible exploitation of these flexibilities and what is economical value that can be captured.
The goal of this internship is to develop a Online Flexibility Audit application that
gather in a interactive and user-friendly way the key information needed to assess the flexibility potential of the industrial plants
compute in a systematic way the savings that could be generated on the different electricity markets
determine the modules of ENERTOP, N-SIDE Energy Flexibility Optimization tool, that are required to model the flexibility, and their associated parameters.
As an intern, your objective will be to design and implement such application before applying it to different use cases in various industries.
The internship positions are based in Louvain-la-Neuve, Belgium
Duration : min 2 to 3 months
Interested to join a young and dynamic team ? Apply through www.n-side.com