snip.
← Back to Home ID: 44195
+Lab @ University of Bristol @pluslab.bsky.social
Jul 11, 03:36 AM

And now for a fiendish (unresolved) question from @martinfleischmann.net: "What is a building?" Sometimes questions with apparently obvious answers are very confusing on closer inspection!

🎤 Whisper Transcript (en) ⏱ 170s

"I was very pleased seeing Sean's first slide of the lightning talks with the question, what is urban? This very nicely rhymes with mine, what is a building? And this is a question which I feel is slightly abandoned. And I think the main reason is that until we start thinking about the answer, the answer seems obvious. As a result, we're not even asking the question. But as soon as we start thinking about the answer, it becomes way more complicated. We have to start questioning ourselves where the boundary of a building should be. Let's take a city like Bristol. If we are in the suburbs, it's easy, typically. There's one detached family house. That's going to be a single building. What if there's a duplex, two of them joined together? Is it one building or two? What if there's a row house structure? How many buildings do we have there? Then we can move into the city centre and we see nicely formed urban blocks composed of eight buildings, for example. Are those individual buildings or not? What if they stand on the same underground parking? Is it one building or is it eight buildings? It's a question of boundaries. And we can also look at this using different dimensions. When the structure becomes a building, is your garden shed a building? Is it not? How does it need to look like to be considered a building? If you say, no, it's not a building, let's imagine your garden shed and let's move to Global South, where a lot of informal settlements is just a huge amount of garden sheds and people lived in them all the time. Do they live in buildings or not? I think we need to ask this question every time we work with building data. And unfortunately, I've seen that people are not doing so, even people who are supposedly releasing large building footprint data. As a result, you can claim there's a data set with 2.5 billion building footprints, claiming we built all of them, we mapped all of it around the globe, but that's not true. You have 2.5 billion polygons using different definitions depending where you are. It's not comparable. It's not something we can rely on. And it's not something we can use to answer many of the questions which were posited here before me. So I think that we should come to the very basics of definitions and start asking the fundamental questions about the components which are forming cities like what is a building. Thank you."

💬 Discussion

+Lab @ University of Bristol @pluslab.bsky.social · Jul 9, 11:39 AM

And now for a fiendish (unresolved) question from @martinfleischmann.net: "What is a building?" Sometimes questions with apparently obvious answers are very confusing on closer inspection!