Trace
First buildAgricultureTrace is better visibility in agriculture, starting with livestock, in a country where the infrastructure already exists: every Norwegian sheep and goat carries an electronic ISO ear tag within 30 days of birth, by law. Three things are built: a tag reader that speaks the open standard those tags already use, a trace system that binds every event to an animal's identity, and a weight cell that ties every weight to a tag. The work now is binding it all together, and researching, area by area, what equipment actually answers the visibility question.
Work in progress, in the open: this page grows as the research and the builds do. Norwegian version: /no/trace.
The tags are already on the animals
Roughly 885,000 winter-fed sheep in Norway are walking RFID infrastructure: ISO 11784/11785 transponders, mandated by Mattilsynet, supplied by exactly three approved makers. The identity layer is done and paid for. What costs money today is reading it, weighing against it, and keeping the history, and that is exactly the part that can be open.
The one metric that makes visibility real
A healthy Norwegian lamb should gain about 300 grams of live weight per day; lambs growing under 40 grams per day are a documented welfare problem. The difference between those two animals is invisible to the eye until far too late; it is only visible in repeated weighings tied to the same identity. That is the whole argument in one sentence: the tag gives the animal a name, the reader captures it in the race, the weight cell attaches the number, and Trace keeps the curve. Growth per day, per animal, over time: the metric the entire meat production economy already runs on, currently rationed by equipment prices.
One ewe lamb, fully visible
What Trace shows for a single animal: identity from the tag, every event bound to it, and the growth curve that makes health visible weeks before the eye can. Simulated data: the devices are on the bench.
Weight in kg · the dashed line is the ~300 g/day healthy-growth reference; under 40 g/day is a documented welfare flag · hover or tab the points · the table below has every reading
| Date | Weight | g/day |
|---|---|---|
| 2026-04-15 | 4.8 kg | born |
| 2026-05-15 | 14.2 kg | +313 |
| 2026-06-15 | 23.8 kg | +310 |
| 2026-07-15 | 32.9 kg | +303 |
| 2026-07-29 | 37.3 kg | +314 |
The curve shows the shape; the table is the record. Growth per day is computed between consecutive weighings: 310 g/day across the whole 105 days. Simulated data, one tag identity.
Area by area: what actually answers the visibility question
Trace researches equipment and technology the same way Manufacturing researches its industry: sourced, priced, and honest about what is commodity and what is locked. First pass, recorded 2026-07-29.
Satellite crop vigor is a solved commodity: Sentinel-2 gives free 10-meter NDVI through Copernicus, and Norwegian services layer free biomass maps on top of it. Weather is free too: NIBIO runs over seventy agricultural stations with open data. The honest limitation at Norwegian latitudes is cloud cover, not access or price. Trace links to these; it does not rebuild them.
Norway's actual livestock trackers run on Telenor's nationwide NB-IoT and LTE-M, not LoRaWAN, which follows the fiber footprint into towns and stops before the utmark. Trace hardware is therefore designed for NB-IoT, LTE-M, or local WiFi with store-and-forward. A dependency named, like always.
Norway's agricultural data sits in cooperative-owned silos: Animalia's Sauekontrollen and Storfekjøttkontrollen, TINE and Mimiro's Eana, Nortura's Min side, with Landbrukets Dataflyt as the settlement-document pipe. Every actor exchanges data through bilateral agreements; none exposes a farmer-facing API. Two signs the wall is cracking: Mimiro open-sources its entire data-hub technology, and Mattilsynet is building an open-source production-animal API in public. Trace's position is to be ready on the open side of that wall, capturing cleanly today, connecting the moment the registers open.
What is, and is not, claimed
The three devices are first builds by the founder: bench-proven, not season-proven; this page will track the field record honestly, including what breaks. An open scale cannot be used for economic settlement without Justervesenet conformity assessment. Trace weighing is a management tool, and settlement weights come from the slaughter line anyway. Direct sync into Sauekontrollen requires either Animalia's device list or manual import of Trace's exports (a partnership question, stated plainly rather than promised away). And tag manufacturing is not on the table: ICAR conformance is a manufacturer-scale process, and Norway's three approved suppliers do that job well. Trace reads what exists; it does not reinvent it.
Where these numbers came from
Mattilsynet: ear-marking rules, sheep and goats ·Mattilsynet: approved tag suppliers ·Lovdata: landdyrsporbarhetsforskriften ·ISO 11784/11785: the open ID standard ·FDX-B protocol: bit-level documentation ·No reader approval required: regulator statement ·OS ID: commercial reader pricing ·ICAR: certification schemes (none for scales) ·ICAR: beef recording, scale accuracy guideline ·NSG: 300 g/day growth and spring weighing ·NVT: low slaughter weights as a welfare problem ·SSB: the Norwegian sheep population ·Sauekontrollen member agreement: data ownership ·Mimiro: open-source data hub ·Mattilsynet: open production-animal API, in progress ·Justervesenet: settlement-weighing rules ·Telenor: NB-IoT/LTE-M coverage ·NIBIO: open agricultural weather stations ·Copernicus: free Sentinel-2 data