Customer
Tipspeed was founded in France in 2023. Their goal is simple: to help wind farm operators ensure that their turbines are converting all available wind into energy.
Tipspeed accomplishes this by combining existing data with physics-based models to recreate the wind each turbine received, and calculate the expected production. By comparing this with the actual output, they find discrepancies and identify ways to improve performance.
So far, Tipspeed has analyzed over 2 gigawatts across Europe, the US and India, and their clients have already recovered 31 gigawatt hours of annual energy production—roughly equivalent to adding three modern turbines, all without any extra hardware.
Challenge
Wind farm operators need to compare the electricity produced by each turbine with the wind a turbine actually receives. However, wind levels vary hugely, and operators rarely have a clear, independent measure of it as the standard nacelle anemometer (the sensor used to measure wind speed) is far from reliable. So when production is lower than expected, it’s not always easy for operators to say if their turbines are having issues, or if there was simply less wind than expected.
Turbines constantly rotate to face the wind, and their orientation is tracked by sensors on the nacelle. But this alignment is never perfect: due to local wind conditions, sensor maintenance, or software issues, turbines can end up pointing off the wind. This misalignment leads to loss of energy and possibly even structural damage.
Tipspeed models wind direction independently to compare this with where turbines are pointing. Satellite imagery gave them a way to measure turbine orientation without any on-site personnel or equipment. However, they needed very high resolution imagery to pinpoint turbine direction, and they also needed multiple images of each turbine over each analysis period.
Tipspeed uses satellite imagery to measure the true orientation of each turbine independently—as the sensors measuring nacelle heading (its bearing relative to north) can themselves be off—and compares it with their own wind direction model to detect and quantify this misalignment, without any on-site personnel or equipment.
While they first started working with one provider, they quickly realized that in order to scale up, they’d need much more imagery over larger areas. Opening and managing accounts with multiple providers would be extremely time-consuming, slowing down project deliveries and increasing operational costs. In addition to this, satellite imagery comes in many different formats; standardizing the data would eat up valuable engineering time.
“Working with UP42 gave a huge boost to our satellite imagery workflows for wind farm performance analysis. Besides the tools and services, I also appreciate the interactions with the team. They’re always ready to provide us support when needed.” —Samuel Davoust, Co-Founder & Science Lead at Tipspeed
“For me, what matters is being able to serve our clients anywhere in the world. UP42 gives us access to the satellite imagery that makes this possible, at the scale we need." —Florian Rebeyrat, Co-Founder & Market Lead at Tipspeed
Solution
Tipspeed first encountered UP42 while searching for additional imagery vendors. To them, the most immediate advantage was the variety of providers available on the UP42 platform. The platform makes it easy to search for and order data according to their needs, and the huge breadth of sensors ensures that they’re able to find images for any wind farm they need to monitor. By unifying search and ordering into one interface, Tipspeed eliminates the need to set up and manage separate contracts and workflows for each provider.
Tipspeed uses the browser-based console to explore imagery options and manage orders. They also automate data retrieval using UP42’s Python SDK, downloading imagery directly into each specific project, enabling faster project turnaround. Once they have the imagery, they process it, identifying individual wind turbines and their key components to derive their directions.
Imagery metadata, such as satellite azimuth and elevation angle, also provides key information. Because the UP42 platform standardizes all delivered data, Tipspeed doesn’t need to spend time aligning data from different providers. This standardization is another value add for the platform, as it frees up engineering resources for other critical tasks, and allows Tipspeed to confidently take on new projects all over the world, scaling without increasing operational overhead.
Result
Since they started working with UP42, Tipspeed has seen a broad increase in geographical coverage and the ability to scale up projects. They’ve also saved time on searching for imagery from different providers, as well as on properly formatting files and metadata.
Prior to UP42, each new provider took a few days of software development to integrate into their workflows. Tipspeed has already saved over two weeks of software development thanks to the UP42 platform’s standardized approach, with much more time savings ahead. They also estimate that the platform has allowed them to take on 50% more projects than before, while keeping the same level of quality.
In one specific case, a client had identified misalignment on a few turbines using LiDAR (both ground-based and turbine-mounted), but needed to check the entire site (over 50 turbines). Doing LiDAR scanning on each turbine would take a long time, and would also be very expensive.
Instead, Tipspeed used satellite imagery from UP42 and their physics-based wind modelling to measure misalignment on all turbines. The client applied corrections on a large number of turbines, and confirmed that they’d seen a production increase of 1%, a massive gain in an industry where the typical project internal rate of return is between 7-10%.
A processed image showing nacelle direction (the direction the turbine is facing) versus the wind direction
Tipspeed is continuously working to improve the quality and automation of their processes, and they see UP42 as a key partner in this. In the months ahead, they’re planning on using UP42’s Python SDK to further automate image retrieval, granting continuous monitoring to their clients to ensure that turbines are performing as expected.