banner
LightNews
robintw.bsky.social
Robin Wilson
@robintw.bsky.social
500 followers 390 following 110 posts
Freelance geospatial software engineer and data scientist. PhD in satellite imaging. Blogger (https://blog.rtwilson.com/) author of the FreeGISData site (https://freegisdata.rtwilson.com/) & the British Placename Mapper (https://placenames.rtwilson.com/)
blog.rtwilson.com
Posts Media Videos Starter Packs
robintw.bsky.social
Robin Wilson @robintw.bsky.social · 9d
Very proud of my former student (now colleague!), Gabi, who just gave her first conference presentation at #foss4guk in Leeds.

#geospatial #gis #osgeo #postgis
1 3
robintw.bsky.social
Robin Wilson @robintw.bsky.social · 10d
The new 'gdal raster neighbours' function is pretty cool - it does convolution across a raster, with configurable kernels and other options. Not yet released, but available in dev builds. See gdal.org/en/latest/pr... for details.
gdal raster neighbors — GDAL documentation
gdal.org
1 1 4
robintw.bsky.social
Robin Wilson @robintw.bsky.social · 11d
Looking forward to #foss4guk this week in Leeds.

I'll be attending with a colleague, running a workshop on producing live vector tiles from PostGIS and my colleague will be giving a talk.

Now just to hope that the trains work better than last year: I got stuck in Bristol TM because the lift broke!
2
robintw.bsky.social
Robin Wilson @robintw.bsky.social · 12d
Debated about buying some heavy water today...for my #emfcamp talk/demo but a) it's very expensive and b) it'd probably get me put on a list. I'm consoling myself by buying high wattage halogen lamps instead.
Reposted by Robin Wilson
sqlliz.bsky.social
Elizabeth Garrett Christensen @sqlliz.bsky.social · 23d
@pwramsey.bsky.social and I are hosting PostGIS Day again this year! The call for papers and registration is open now. Join us!

www.snowflake.com/postgis-day-...
PostGIS Day | November 20
Learn how others are making use of PostGIS, pick up some tips and tricks, and share stories about bringing this excellent tool into your organizations.
www.snowflake.com
5 7
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Sep 7
This is one of the weirder comments that I've made on BlueSky, but...we have that Z80 chopping board!
1 1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 23
(Partially prompted by reading Exploding the Phone, a good book about phone phreaking in the US, and realising that from that and various other similar books - eg. Kevin Mitnick's - I know significantly more about the US phone network than the network in my own country!)
1 1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 23
Does anyone know any good books about the UK telephone network? Ideally I'd be interested in a history of the network, exchange development, memories of working on the network, big events (eg. phONEday etc) - but I can't seem to find anything like this
1
Reposted by Robin Wilson
eric.ness.net
Eric Ness @eric.ness.net · Aug 8
Orbio Earth just open-sourced their full methane detection pipeline—models, tools, synthetic-plume engine & validation notebooks—for Sentinel-2, Landsat 8/9, and EMIT. #Methane #ClimateTech #OpenSource - orbio-earth.github.io/Project-Euca...
3 11
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 7
I'm also quite surprised that I hadn't come across this previously, having been working in geospatial for quite a long time. Every day is a day for learning!
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 7
So, TIL that a MultiPolygon is invalid if any of its polygon parts touch at more than a single point. Eg. a MultiPolygon where two 'polygon parts' share an edge is invalid.
I can see why this is the case - those two polygon parts should be merged, but it's a bit annoying.

#gis
1 3
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 5
I won't be able to make this as my wife is in the US at the time and so I'm needed at home - but I've had a great time at PyCon UK in the past, and I'm already looking forward to watching the videos of some of these talks.
pyconuk.org
PyCon UK @pyconuk.org · Aug 5
📣📣 Announcement: full PyCon UK 2025 programme is now LIVE! 📣📣

It’s here, folks! The full schedule is up, and it’s chock-full of talks, workshops & more! 🎉🐍🌟

Schedule is over here 👉 buff.ly/9UtKEOl
Tickets this way 👉 buff.ly/83NpL2R

#Python #TechEvents #development #manchester #PythonProgramming
Schedule – PyCon UK 2025
PyCon UK Friday 19th September to Monday 22nd September 2025, Contact Theatre, Manchester
2025.pyconuk.org
1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 5
Gabi, a MSc student I've been supervising, will also be giving a talk about developing a system to look at secondary flooding impacts, specifically loss of power and loss of road access, on businesses in the UK. She'll be covering use of pgRouting, complex PostGIS SQL queries etc. #foss4guk
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 5
Pleased to announce that I'll be giving a workshop at #foss4guk entitled 'Live Vector Tiles from PostGIS with pg_tileserv: A Hands-on Workshop' - I'll take you through the full process of using pg_tileserv to serve vector tiles to web maps (using MapLibre, Leaflet etc) and QGIS. Come along!
1 1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
I might have to steal one of their section headers as my tagline, as I tend to refer to myself as a geospatial software engineer: 'Geospatial software engineering as its own weirdo niche'

#gis
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
The Placing Technology blog (placing.technology) is utterly fascinating - someone blogging parts of their PhD research about the history of GIS and geospatial technologies in general.
Placing Technologies
The latest entries posted on Placing Technologies
placing.technology
1 1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
I don't normally write long threads, but was suddenly inspired. Hope that helps @ajgwords.bsky.social - I'll be attending FOSS4G UK (and presenting myself), so hopefully I'll be able to hear the talk!
1 2
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
It needs to have changed a lot from my days as an undergraduate (~20 years ago - ugh!), but I suspect it either hasn't changed that much, or the sort of things you can do in an undergrad practical session don't let you explore what is *really* done in RS any more.
1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
now with cloud-based systems, STAC and so on. I'm not sure how we should deal with this as a community, and I'm not sure how modern methods will 'trickle down' to non-code based environments. Like, what do you do in an undergrad RS practical session these days? 8/n
1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
4) There's a big gulf between what is possible with remote sensing if you can write code, and what is possible if you can't. That's fine for me - I can write code and enjoy doing so - but not everyone can or wants to, and that limits what you can do hugely. Potentially that applies even more 7/n
1 1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
3) Practical remote sensing is often taught poorly, and often taught using inappropriate software. It's a long time since I've taught remote sensing at uni level, but back then and still now I keep coming across massively silly ways of teaching you how to actually do stuff with RS data 6/n
1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
2) Loads of methods are developed using very high resolution data which can be a problem because the data is *really* expensive (sometimes you can get it free in academia), making those methods too expensive to use in the real world. Often high-res data isn't actually needed - it's 'too high' 5/n
1 2
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
much more that can be done with remote sensing than that.
That's my biggest bugbear, others are:
1) Far fewer people care about calibration and atmospheric correction now, as they don't normally need to worry about it. That's mostly ok, but *someone* needs to worry about it or it breaks down 4/n
1 1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
or moisture content of vegetation, actually measuring air pollution etc. I know some ML methods are used to try and estimate these physical quantities, but so much stuff these days seems to be about just finding objects in satellite images (roads, cars, swimming pools, whatever) and there is so 3/n
1
robintw.bsky.social
Robin Wilson @robintw.bsky.social · Aug 4
with what seems like little understanding of the physical processes and science behind what is actually going on. I give talks about remote sensing to public groups (science clubs, U3A etc) and they are often amazed at how we can do *actual science* from orbit, measuring chlorophyll content 2/n
1