Comment by mike_hearn
4 days ago
I wonder if they've validated demand with the water companies. Their website doesn't seem to talk about any successful deployments or partnerships with industry, beyond something that was getting started just as they were shutting down and which isn't robot related.
Thames Water already uses pretty hi-tech approaches to finding leaks including using underground microphones to find them via sound (with obligatory AI, of course), and then they have been trialling a chemical that can travel through the pipes and jam the leaks.
https://www.thameswater.co.uk/always-fixing/finding-leaks-wi...
https://www.aquatechtrade.com/news/urban-water/thames-water-...
https://en.gutermann-water.com/2019/09/23/thames-water-finds...
They also know all about pigs, of course. So I wonder what the researchers felt their edge was that would let their approach do better than others. It seems from a quick check like the water companies are still primarily interested in hydrophonics and how to adapt those to plastic pipes.
About a decade ago I worked at a water consultancy company (of which Thames Water happened to be a client) to build a catalogue website for pipe condition assessment technologies. Already at the time that I was building it, there were dozens of these finished products with all kinds of sensors for many sizes of pipes, be they dry or with water still flowing inside. Yes, there is some demand for these devices, but there is also a lot of technology out there already for many use cases.
All that to say: I too wonder what makes this one so special.