Post
102
We're releasing a free radar dataset aimed at ML practitioners who want to work with radar data but don't have access to the hardware or domain expertise to collect their own.
Why this exists
Radar is underrepresented in the ML community relative to its potential. If you're working on sensor fusion, object detection, tracking, or signal processing, you've probably noticed that finding realistic, well-documented radar data is harder than it should be. Most of what's out there is either synthetic, poorly labelled, or locked behind defence contracts.
We're a UK radar design house (Plextek) and we've been building radar systems for 35+ years. We have the hardware, the test environments, and the RF engineering knowledge to produce quality datasets. What we don't have is unlimited ML talent in-house, which is exactly why we want to get this data into the hands of people who do.
What you get
Curated real-world radar captures for non-commercial experimentation
Domain support from our radar engineers if you need help interpreting the data (radar data can be unintuitive if you haven't worked with it before)
A route to collaboration if your work leads somewhere interesting
Who this is for
Anyone training models, benchmarking algorithms, or just curious about radar as a sensing modality. Startups building sensor fusion pipelines. Research groups. Hobbyists. If you've wanted to play with real radar data but didn't know where to start, this is the on-ramp.
Access the dataset
Request access and get full details here: plextek.com/learn/radar-data-for-experimentation
Happy to answer questions about the data, radar fundamentals, or what kinds of ML problems radar data is well suited to.
Why this exists
Radar is underrepresented in the ML community relative to its potential. If you're working on sensor fusion, object detection, tracking, or signal processing, you've probably noticed that finding realistic, well-documented radar data is harder than it should be. Most of what's out there is either synthetic, poorly labelled, or locked behind defence contracts.
We're a UK radar design house (Plextek) and we've been building radar systems for 35+ years. We have the hardware, the test environments, and the RF engineering knowledge to produce quality datasets. What we don't have is unlimited ML talent in-house, which is exactly why we want to get this data into the hands of people who do.
What you get
Curated real-world radar captures for non-commercial experimentation
Domain support from our radar engineers if you need help interpreting the data (radar data can be unintuitive if you haven't worked with it before)
A route to collaboration if your work leads somewhere interesting
Who this is for
Anyone training models, benchmarking algorithms, or just curious about radar as a sensing modality. Startups building sensor fusion pipelines. Research groups. Hobbyists. If you've wanted to play with real radar data but didn't know where to start, this is the on-ramp.
Access the dataset
Request access and get full details here: plextek.com/learn/radar-data-for-experimentation
Happy to answer questions about the data, radar fundamentals, or what kinds of ML problems radar data is well suited to.