iiinova — where your digital self keeps evolving

Michael Szell @mszll.datasci.social.ap.brid.gy · 19/10/2026

Next semester I'm teaching the course in #Geospatial Data Science at ITU #Copenhagen. I'm looking for a talented TA to run the exercises. This role is perfect if you are currently studying, or have just finished your studies in Data Science, or related fields […] [Original post on datasci.social]

0615 15

Michael Szell @mszll.datasci.social.ap.brid.gy · 08/10/2026

The first hands-on Python session for #BikeNetKit, today at #D3A, went well! Our 15 min tutorial for GrowBikeNet is also available online - try it out here: github.com/BikeNetKit/tutorials/blo… Here is how our software would fix Nyborg's bike network:

Map showing Nyborg's existing protected bike network in purple and in-between pieces in green built by GrowBikeNet

047 7

Michael Szell @mszll.datasci.social.ap.brid.gy · 08/10/2026

We had a full room with a very diverse crowd at today's #D3A session on human-centric mobility! The invited talk by Carlos M. Lima Azevedo on Emotional Cities was a nice bridge between the gender mobility and cycling parts.

Room with around 40 people listening to a speaker

Speaker in front of slide titled Gendered Mobility

Speaker in front of slide showing Paris' bicycle network

032 3

Michael Szell @mszll.datasci.social.ap.brid.gy · 08/10/2026

At #D3A, come to Salon 25 at 11:00 for Data Science of Human-centric Mobility! 🚶 🚲 🚋 d3aconference.dk/data-science-of-hu…

Info-Flyer for the D3A parallel session Data Science of Human-centric Mobility showing the faces of the 4 organizers and text:

This 90-minute workshop will bring together professionals from data science and network science who research human-centric mobility and who develop software applications, and everyone else who is interested in the topic and its applications, especially in the context of the green transformation in Denmark.

The workshop will feature few short, engaging scientific talks on the topics of human mobility and bicycle networks, an interactive hands-on session to introduce the open-source Python software suite BikeNetKit with interactive visualizations, and discussion sessions in-between, for an audience to engage in diverse ways with the state-of-the-art of the field.

052 5

Michael Szell @mszll.datasci.social.ap.brid.gy · 22/09/2026

#CS2NORDICS is over. Fantastic talks - Fantastic conference! Here photos from talks on the #bicycle #network topic.

043 4

Michael Szell @mszll.datasci.social.ap.brid.gy · 21/09/2026

Today's & tomorrow's #CS2NORDICS in Copenhagen will feature 3 presentations from our #bicycle #network research: nosocss.org/program.html 1) Manuel Knepper: #BikeNetKit: Open-source bicycle network analysis 2) Riccardo Basilone: Cyclicity: a […] [Original post on datasci.social]

BikeNetKit banner showing a network and 3 approaches to improving bicycle networks

075 7

Michael Szell @mszll.datasci.social.ap.brid.gy · 16/09/2026

Nice paper, finally on arxiv: Higher-order interactions reveal synergistic backbones of cycling infrastructure networks arxiv.org/abs/2609.16285

Street network of Hamburg with a "nicely" grown bicycle network of 65 km that looks like a ring road. Starting from the set of the ten bidirectional links with the highest length-normalized first-order importance (orange), we iteratively add links with the largest length-normalized cross-synergy (see Methods for details). The upgraded links (blue) form a connected network backbone already for a small total length of 65 km of added bike paths, maximizing the contributions of cross-synergies to total network performance.

051 5

Michael Szell @mszll.datasci.social.ap.brid.gy · 14/09/2026

Our research assistant Manuel Knepper is soon on his way to #CRBAM to present this Friday morning: #BikeNetKit : Developing an open-source bicycle network software suite for urban planners pretalx.com/crbam-2026/talk/QAZLGF

Title slide titled "BikeNetKit: Developing an open-source bicycle network software suite for urban planners"

046 6

Michael Szell @mszll.datasci.social.ap.brid.gy · 14/09/2026

For the 22nd birthday of our free #MMOG #Pardus, we implemented a new NPC, the Cyborg Corebreaker, and new advanced skills. ...we are back after some years of development pause. 😅 www.pardus.at/index.php?section=arc…

The Cyborg Corebreaker is a massive new Cyborg vessel, built around a heavily reinforced core and designed to survive punishment that would tear lesser constructs apart. Its imposing manta-like frame combines thick armour and powerful weapon systems. Pilots looking for a new challenge should prepare accordingly. The Corebreaker was not built to be dismantled easily.

001 1

Michael Szell @mszll.datasci.social.ap.brid.gy · 07/09/2026

Come to #D3A @ Nyborg on Oct 8 - We are organizing the workshop Data Science of Human-centric Mobility! d3aconference.dk/data-science-of-hu… Showcasing latest research and innovations like #BikeNetKit!

153 5

Michael Szell @mszll.datasci.social.ap.brid.gy · 30/08/2026

Due to a >1000% increase in attempted signups on our server datasci.social in the past weeks generated by AI bots, we added a new rule: Only human users. (They won't care, but it's now explicit.) We also clarify our ads rule: No commercial advertisements […] [Original post on datasci.social]

Screenshot of the server rules' git diff, showing green and red lines of deletions/additions.

021 2

Michael Szell @mszll.datasci.social.ap.brid.gy · 05/08/2026

City2Graph: A Python library for Heterogeneous Graph Neural Networks and spatial analysis in urban systems www.sciencedirect.com/science/artic… Looks very interesting!

Conceptual diagram of the City2Graph framework.

0125 12

Michael Szell @mszll.datasci.social.ap.brid.gy · 31/07/2026

Very nice new tool showing #bike routes in #Austria with lots of configurability: radatlas.at There is a family-friendly option, but there isn't more than a handful of routes suited for families with small children. In other countries you have a real […] [Original post on datasci.social]

075 7

Michael Szell @mszll.datasci.social.ap.brid.gy · 27/07/2026

It seems #Google #scholar has now really become #slopware. Every week I get weirder and weirder edit suggestions, this time to revert several published papers to their preprint versions.. Is #openalex the future of scientometric data/search? 🫣

Google scholar edit suggestion to revert several published papers to their preprint versions

033 3

Michael Szell @mszll.datasci.social.ap.brid.gy · 24/07/2026

Invisible walls: how pedestrians navigate around social interactions arxiv.org/abs/2607.20876

A diagram of our field experiments. Actors (shown in orange) stood in busy pathways while
displaying 5 distinct signs of interactional involvement (see depictions of Variables 1–5). In different configurations of these variables, we observed whether pedestrians (shown in blue) walked between or around our
actors and we measured pedestrians’ walking trajectories.

012 2

Michael Szell @mszll.datasci.social.ap.brid.gy · 23/07/2026

It's nice to work hard on things in silence, and imagine they could become really useful someday.. docs.bikenetkit.org/GrowBikeNet/usa… #BikeNetKit

Street network of Athens with its primary and secondary roads highlighted in red

052 5

Michael Szell @mszll.datasci.social.ap.brid.gy · 17/07/2026

A happy, explorative trip in the sun! 🌞 🌼 (Only interrupted every 40m by a high-risk, life-or-death situation, where anytime an #SUV could have exited and crushed her. 🤷 ) #urbanplanning #parenting #motonormativity #CIM #Copenhagen

Small child in green dress kneeling to smell flowers in front of row of parked SUVs.

Small child in green dress kneeling to explore a blue door in a tree.

Small child in green dress riding tricycle on sidewalk, next to an SUV parking in driveway.

A small child in green dress riding a tricycle on sidewalk, with low-vision driveway approaching on the left from which an SUV could exit anytime.

021 2

Michael Szell @mszll.datasci.social.ap.brid.gy · 16/07/2026

The slides of my talk two days ago at #Complexity Science Hub #Vienna are now online [pdf]: michael.szell.net/downloads/talk_sz… The complex socio-technical system of #bicycle networks: From politics to planning software […] [Original post on datasci.social]

Title slide showing a bicycle network in Vienna and the title: The complex socio-technical system of bicycle networks: From politics to planning software

0107 10

Michael Szell @mszll.datasci.social.ap.brid.gy · 16/07/2026

Every time I'm in #Vienna I can't resist the urge to an impulse purchase at Doblinger. This time it hit on the more mysterious side.. 👻

Photo of two Urtext versions of Skrjabin's Preludes (Op 11) and Satie's Nocturnes, on top of piano keys

021 2

Michael Szell @mszll.datasci.social.ap.brid.gy · 10/07/2026

Tuesday, July 14th, 2pm I'll give a talk at Complexity Science Hub #vienna : The complex socio-technical system of #bicycle networks: From politics to planning software csh.ac.at/events/the-complex-socio-…

3104 10

Michael Szell @mszll.datasci.social.ap.brid.gy · 08/07/2026

Trysalin Nelson opening #agit2026! First time for me here - happy to see a strong #geospatial community in Austria.

Speaker and moderator in front of screen

021 2

Michael Szell @mszll.datasci.social.ap.brid.gy · 08/07/2026

We find a strong correlation between right wing votes and delays in bike network development in #paris. Qualitative observations for such a link have been reported in the past, but we now have the data to back it up.

Left: Scatterplot of the share of bicycle lanes built
against left-wing votes
, Right: against right-wing
votes, with corresponding maps of the voting patterns. The black lines show fitted linear regressions. Dashed gray lines show the average values for each axis.

46921 69

Michael Szell @mszll.datasci.social.ap.brid.gy · 08/07/2026

🚨 New paper on #cycling #politics: Right-wing Votes Relate to Delays in Bicycle Network Development findingspress.org/article/163938-ri…

Left: Paris’ Plan Vélo 2021-2026 progress on street level, Right: share of bicycle lanes built
aggregated by political voting unit

23720 37

Michael Szell @mszll.datasci.social.ap.brid.gy · 04/07/2026

Looking forward to #agit2026, presenting our analysis on the Danish Bicycle Node Network, and our progress on @bikenetkit bikenetkit.org #BikeNetKit July 9th, session 10B from 13:00. See you in Salzburg!

Title slide with title, authors, and a colorful map/network of Denmark

052 5

Michael Szell @mszll.datasci.social.ap.brid.gy · 10/06/2026

TFW you invest a whole day of work into something, unsure if it's worth the effort, and then you see the outcome 🤩 #quadrangulation

Manhattan street network, quadrangulated

284 8

Michael Szell @mszll.datasci.social.ap.brid.gy · 09/04/2026

Now we developed the loop census method, which lists each node's loops (cycles) up to day-trip length. We find that long-range cyclists can access most of the country with many choices, but families with small children can not - which could however be overcome by e-bikes.

Two maps of Denmark side by side showing the bicycle node network for adult cyclists (left) versus families with children (right). 38% of the nodes in the adult cyclist scenario are red (no round trip options), while 79% of nodes are red for the family scenario.

131 3

Michael Szell @mszll.datasci.social.ap.brid.gy · 09/04/2026

A bicycle node network has a wayfinding system of numbered nodes for recreational #cycling, on a regional/national scale. We have previously shown on bikenetwork.dk the low bikeability in #Denmark outside of cities, now improved by the new bicycle node network.

A signpost with a node number and two arrows pointing to adjacent nodes

The Danish bicycle node network in green, spanning Denmark (except greater Copenhagen)

Map of Denmark showing high bikeability in cities (green), low outside of cities (red)

101 1

Michael Szell @mszll.datasci.social.ap.brid.gy · 09/04/2026

New 🚲 preprint! arxiv.org/abs/2604.07029 Quality assessment of a country-wide bicycle node network with loop census analysis We analyzed the new 30000km long Danish #bicycle node network, exploring all 28M round trips in the network up to day-trip length. 🧵

Network of 7000 nodes and 10000 links spanning Denmark, nodes colored from red to blue.

12416 24

Michael Szell @mszll.datasci.social.ap.brid.gy · 27/02/2026

Just registered for #gisruk, woohoo! Who else is coming to Birmingham? I'll talk about #bicycle node networks.

Map of the Danish bicycle node network, showing many colored nodes over the country

053 5

Michael Szell @mszll.datasci.social.ap.brid.gy · 12/01/2026

Mind the gap: Revealing sidewalk networks at scale pdf: iopscience.iop.org/article/10.1088/… Nice to see important research on the both most important and neglected mode of transport 🚶

Figure 2. Schematic pipeline of the network construction process. The deep learning stage (a-c)
begins with (a) tiling high-resolution aerial imagery (5000 × 5000 pixels) into smaller patches (512 × 512
pixels), which are fed to (b) a segmentation model (DeepLabV3). This model generates (c) binary output
masks identifying sidewalk pixels. The masks are then reassembled to reconstruct the full segmented tile. In
the network construction stage (d-f) the merged predictions undergo a (d) binary closing to fill gaps, (e)
skeletonization to reduce the sidewalk predictions to their centerlines, and (f) generation of the network (nodes
and edges). Finally, these tile-level results are integrated to produce a large network across the entire urban
area

045 5

Michael Szell @mszll.datasci.social.ap.brid.gy · 09/01/2026

Haven't seen this on Mastodon yet - the #royalsociety #musk saga continues: www.ft.com/content/088b804f-15e0-4a… Perfect example for how to run a scientific society into the ground 🤦

Image of an old dude with white hair leaning against royal bookshelf. Next to it the headline:
Elon Musk should keep UK Royal Society membership, says president
Paul Nurse tells FT that national science academy should avoid ‘making judgments’ about ‘character’ of fellows

022 2

Michael Szell @mszll.datasci.social.ap.brid.gy · 09/01/2026

Not the "new year's resolution" I wanted to make 😢.. let's really hope this gets better in a few years:

Since 2026 I have decided to stop reviewing for US-based conferences, so please do not contact me to review for such conferences:

Decline message US-based conferences:
Dear Colleague,
thank you for asking. As much as I would like to accept your invitation, this year I cannot review for the conference unfortunately, as I couldn't attend even in principle because of political repression in the host country. I won't review for conferences in the US for the foreseeable future - thank you for your understanding.

1103 10

Michael Szell @mszll.datasci.social.ap.brid.gy · 05/01/2026

The #bicycle network field has grown so much that there are now papers out analyzing its citation network, haha 😆 www.sciencedirect.com/science/artic… Fun to see how this has evolved! Our nodes are: orozco, szell, vybornova. Several friends also visible 😍

Co-citation network of the bicycle network literature. Colored nodes and links.

165 6

Michael Szell @mszll.datasci.social.ap.brid.gy · 03/01/2026

Just updated our server's downtimes for 2025 and for all-time. Although datasci.social had a downtime of over 10 hours in 2025, our all-time uptime is still way above 99.9%. 🤓 community.datasci.social/docs/server

Historical uptimes
Year Uptime Average downtime per month
All-time 99.931% 30 minutes
2025 99.876% 54 minutes
2024 99.998% <1 minute
2023 99.908% 40 minutes
2022 99.995% <2 minutes

011 1

Michael Szell @mszll.datasci.social.ap.brid.gy · 22/12/2025

Targeted cooling of urban cycling networks for heat-resilient mobility arxiv.org/abs/2512.11753 Cool 😬 paper on how targeted tree planting can substantially cool streets for #cycling!

Land-cover comparison for Midtown Manhattan under the (a) Base and (b) Tree-Planting scenarios, showing added canopy cover across the hottest 1,000 street segments.

014 4

Michael Szell @mszll.datasci.social.ap.brid.gy · 27/10/2025

Very cool paper out by our collaborator Chris Larkin (et al): Identification of plausible low traffic neighbourhoods using open data www.sciencedirect.com/science/artic… Code: github.com/Froguin99/LTN-Detection

Map of Newcastle Upon Tyne, partitioned into potential low traffic neighborhoods

128 8

Michael Szell @mszll.datasci.social.ap.brid.gy · 17/10/2025

Here the city's press release: via.ritzau.dk/pressemeddelelse/1456… They ~quadrupled the bicycle budget to >600 mio. DKK. Below my full comments to the journalist, which was used for the article.

Kilde: Mobilitetsredegørelsen 2025

På cykelområdet er der ca. blevet bevilliget 160 mio. kr. årligt (10-års gennemsnit for 2016-2025). Investeringer på cykelområdet omfatter både midler til cykelinfrastruktur og cykelinitiativer, men også sikre skoleveje, trafiksikkerhed samt eksterne investeringer og bidrag fra puljer. I 2025 blev der bevilliget 174 mio. kr. til cykelområdet. Af de 174 mio. kr. blev 82 mio. kr. bevilliget primær til cykelinfrastruktur og cykelinitiativer, 68 mio. kr. til sikre skoleveje, trafiksikkerhed, strøggader, cykel- og gangbroer og 11 mio. kr. i forbindelse med eksterne investeringer fra trafikselskaber samt bidrag fra puljer.

From an urban planning perspective it is very good and much needed that Copenhagen is increasing its bicycle budget. It is especially important that some of this investment will be allocated to child safety, as Copenhagen's transport policies have indeed neglected children. I personally experience this every morning through the terrifying traffic chaos on the street of my 7 year old daughter's school. As a consequence, Copenhagen has fallen behind several cities in Europe. Investing into cycling is increasing a city's livability, and also comes with massive economic benefits, especially to the public health system.

We should however not forget, that still vehicular infrastructure receives a budget that is many times higher than cycling, so this bicycle budget is only like a drop in the bucket. Despite the laudable efforts, Copenhagen is still missing the political will to boldly transform its transport system towards becoming truly human-centric. Also the public wants this, as one can see clearly in Paris' ongoing success story. So, despite offering fantastic cycling infrastructure for fit adults, Copenhagen unfortunately remains inherently car-centric. A budget increase is great, but not enough by far - it will need a fundamental rethinking of transport, urban living, and crucially the political will to change this.

052 5

Michael Szell @mszll.datasci.social.ap.brid.gy · 17/10/2025

Copenhagen increased its bicycle budget - I was asked to comment: www.verdensmaal.org/nyheder/k%C3%B8…

Three cyclists riding in the foreground, colorful houses of Nyhavn harbor in the background

176 7

Michael Szell @mszll.datasci.social.ap.brid.gy · 06/10/2025

Nice project: A computational framework for quantifying route diversification in road networks arxiv.org/abs/2510.02582 Interactive: divercitymaps.github.io

Overview of DiverCity and global patterns in urban road networks. (a) Example of a trip with low
DiverCity (2.18) in Mumbai. Near-shortest routes significantly overlap, leading to low route diversity. (b) Example of
a trip with high DiverCity (9) in Tokyo, characterized by multiple spatially diverse near-shortest routes. For panels
(a) and (b), inset bar plots show the travel time of each alternative route, with NSRs in blue and non-feasible routes
(exceeding the near-shortest threshold, shown as a dashed line) in red.

0158 15

Michael Szell @mszll.datasci.social.ap.brid.gy · 28/09/2025

Putting cyclists in boxes: An analysis of the bike box www.sciencedirect.com/science/artic… Looks like bike boxes are not very useful, i.e. mostly used like in the image below (which means not being used as intended) - now confirmed by science

Image of a bike box with one cyclists in the top right corner

005 5

Michael Szell @mszll.datasci.social.ap.brid.gy · 15/09/2025

Causal evidence of racial and institutional biases in accessing paywalled articles and scientific data arxiv.org/abs/2509.08299 Nice setup - one more reason why it's a good idea to not put papers behind paywalls or data behind "data available upon request"

Of all references
coming from each country, the proportion of upon-request references (i.e., those made to papers with data available
upon request). Brown and green hues represent countries whose proportion is lower and greater than the median
value, respectively. Countries with fewer than 100 references made to the 91,000 papers in our dataset are excluded
and colored in gray.

Global distribution of the
percentage of references that are made towards paywalled papers. Countries with a lower percentage are colored
brown while countries with higher percentage are colored blue.

013 3

Michael Szell @mszll.datasci.social.ap.brid.gy · 15/09/2025

The Hitchhiker’s Guide to challenges in transport policy research: Towards ANSWERing questions regarding life, mobility, and everything www.sciencedirect.com/science/artic… Looks interesting, and cool paper title! 🤓

Twenty major categories of transportation policy publications.

103 3

Michael Szell @mszll.datasci.social.ap.brid.gy · 08/09/2025

If I was still working on signed networks, I might look into this one 😅

Network of deals and lawsuits with AI companies

031 3

Michael Szell @mszll.datasci.social.ap.brid.gy · 06/09/2025

Very interesting looking paper: In tech we trust: A history of technophilia in the Intergovernmental Panel on Climate Change's (IPCC) climate mitigation expertise www.sciencedirect.com/science/artic… Hopefully someone will also […] [Original post on datasci.social]

Frequency per page of selected terms related to key mitigation strategies in the six full reports of IPCC Working Group III published between 1992 and 2022. The data highlights the overwhelming dominance of the term “technology”, particularly in the third and fourth assessment reports (2001 and 2007), where it peaked at over 1.5 mentions per page. In contrast, demand-side concepts such as “lifestyle”, “behavioural change”, and “sufficiency” remained marginal until a notable increase in the 2022 report, reflecting a shift in discourse. Nevertheless, references to technological change continue to outnumber demand-side options by a factor seven.

122 2