Flexibility simulation & energy-data ingestion
Tools simulating the gains of electrical flexibility control, and a hardware-agnostic sensor-data ingestion system.
Context
Wise.Energy helps companies optimise their electrical assets (solar, EV chargers, heat pumps…) to cut their bill and their CO₂ emissions. I worked on two projects for them.
Simulating the gains of flexibility
Simulation tools, in Shiny (R) and Streamlit (Python), to estimate the gains of electrical flexibility control on heat pumps and other assets (EV chargers, etc.).
- Simulations from a client’s historical data, or from synthetic data generated under their real-world constraints.
- Integration of solar production and market prices (Belpex) via the Elia and ENTSO-E APIs.
- Implemented the control strategies (time-of-use, demand shifting, peak shaving…), building on what already exists rather than reinventing the wheel, and on MILP libraries (mixed-integer linear programming), notably for heat-pump optimisation.
- The goal: quantify, before investing, what the smart control of an asset actually saves.
Ingesting the data, feeding the monitoring
A backend + frontend block to graft onto Wise.Brain, their production app. The bulk of the work: designing the database and the ingestion connectors able to absorb varied sources (manufacturer APIs, database connections…) and funnel them into a shared data model.
- A TimescaleDB time-series database, built to handle quarter-hourly data with ease.
- Hardware-agnostic ingestion, under cost and technical constraints, to stay robust and maintainable.
- At the end of the chain: the app’s monitoring brick, with consumption / production tracking charts for the connected assets and a set of metrics.
Outcome
Simulators that turn raw data into quantified savings, and a reusable ingestion building block underpinning the monitoring of energy assets.