Ground Truth for Maps: What Can Only Be Verified on Site
Satellites, open registries, and crowdsourced edits get a map most of the way there. Here's the remainder that only someone physically present can confirm.
Thoughts on geo data collection, field data, GIS and AI workflow automation — from the team that runs the programmes.
Satellites, open registries, and crowdsourced edits get a map most of the way there. Here's the remainder that only someone physically present can confirm.
Three ways to measure the same customer experience, three different answers. Here's what each method actually catches, and what it's structurally blind to.
An indoor map starts aging the moment the scan finishes. Here's how fast that happens, and what it costs to catch up.
A field guide to the legal basis question every mall operator eventually asks their vendor — and usually asks too late.
Every bridge, retaining wall, and charging station has a file somewhere with its defects logged against a timestamp. Here is what belongs in that file, and the standards that say so.
A working reference for designing ground-truth programs that actually scale: what to sample, how to route it, and how to prove the data is clean.
A working reference for how field teams decide what to visit, in what order, and how to check the result before it ships as data.
A photo is only as credible as the chain connecting it to the visit. Here's how that chain holds together, and where it snaps.
GPS signals die at the front door of every mall and office tower. Here's what actually takes over inside, and which accuracy numbers you can trust.
Three numbers show up on every mall and retail analytics dashboard. They measure three different moments in a shopper's visit, and mixing them up produces bad decisions.
A shelf photo is not a dataset. Here's everything that has to happen between the camera click and a row in a compliance table.
Every address and point of interest has an expiration date. Here's what causes the decay, and what actually slows it down.
The three design decisions that decide whether your mystery visit data is evidence or anecdote — and how AI is changing all three.
A working reference for what goes into an indoor digital twin of a mall or large building, and the field methods used to capture and maintain it.
Phones stopped broadcasting a stable identifier years ago. Here's what mall and store foot-traffic analytics can actually still count — and what it can't.
A single photo shows what something looked like. A dataset proves what it is, where it sits, and whether it still matches the record.
Every map pin is a claim about reality. Here's what it takes to make that claim true, and keep it true.
A field guide to what a mystery visit can prove, how AI is changing the mechanics behind it, and why the same checklist means something different in Berlin than it does in Boston.
Three capture methods, three accuracy profiles, one very different failure mode once you're standing in a glass-fronted atrium at 3am.
Every phone in your pocket is quietly raising its hand. Here's what happens to that signal between your pocket and a mall's quarterly footfall report.
A road-monitoring dataset uploaded this month shows what happens when field photography, GPS, and computer vision converge on infrastructure — and where the method still needs a human check.
A live HERE and Mapillary campaign is paying Europeans to photograph their own cities. The reason why says more about geodata quality than any market report.
Mystery-visit programs built for quarterly PDF reports are being rebuilt around computer-vision verification and continuous checks. Here's what's actually shipping, and what's still marketing.
Munich just staged the industry's biggest reality-capture showcase in years. Malls are the reason it's needed.
A September op-ed on tactile sensor wear explains why touch data goes bad long before anyone notices. The same physics shows up in every camera rig collecting egocentric training data in the field.
Humanoid robot shipments nearly tripled in H1 2026. Three research firms can't agree on what most of those robots are actually doing.
A new industry benchmark shows quality verification became standard practice in insights work this year. Physical AI data collection is still treating it as optional.
A recent operational disclosure from data-infrastructure vendor Unidata shows what egocentric capture looks like once it stops being a research project and becomes a running facility.
A warehouse robotics company just said the quiet part out loud: the mess is the product. Here's why that logic applies far beyond robotics.
A RoboBusiness talk on humanoid motion just named the real bottleneck: hands, hips, and torsos need completely different data. Field research has the same problem.
ESOMAR just formalized rules for AI-generated survey respondents. The accuracy numbers are real, and so are the gaps underneath them.
Hugging Face crossed one million public datasets this week. For teams building physical AI, that milestone measures noise almost as much as it measures progress.
SoftBank just backed a construction robotics company whose machines survey the site while they dig it. That's not a small detail.
Three separate reports published this August put real dollar figures on robot training data for the first time. The numbers explain why data collection just moved from the lab to the field.
A multi-point turning feature grabbed the headline. The real story is buried in the data-throughput number nobody asked about.
A new arXiv dataset turns 204,000 clips of human hands into 17 million affordance labels, without a single human annotator marking a frame.
AI can summarize ten thousand survey responses before lunch. It still can't tell you whether the shelf tag was actually there.
A startup put cameras on a thousand real heads. A new retail dataset just proved that where you point the camera matters more than how much footage you collect.
A dataset announcement from Figure AI, paired with fresh moves from LG, Nvidia, and SEER Robotics, shows the real bottleneck in physical AI isn't models — it's field-grade data infrastructure.
A new synthesis pipeline claims to be the largest ego-to-robot dataset to date. Here's what it actually measured — and what it didn't.
A new arXiv paper on converting human video into robot training data quietly explains why raw field data almost never arrives ready to use.
A headset startup just raised real money to record human hands. The harder problem it's actually solving isn't the camera.
Two funding announcements and a dataset paper from the last month show the same thing: robotics companies are now building physical operations, not just models.
Public egocentric datasets crossed a million hours in 2026, but teams report discarding 90% of bulk footage before training. Why coverage-designed collection is the complement bulk data can't replace.
The EU pushed high-risk AI obligations to December 2027. What the delay does and doesn't change for teams collecting training data now.
What it actually takes to capture first-person training video, and why robotics labs can't get enough of it.