SoilData.ai™ verification for biologicals

A printed biology sensor plus soil and air readings, so commercial teams can see whether living soil responds where product was applied.

How a verification pilot works

One site, matched nodes, one timeline.

Install nodes in treated and untreated areas

Matched treated and untreated nodes at the same site, set during scoping with your agronomy lead.

Read living-soil activity at each node

The biology sensor tracks microbial activity, alongside soil and air readings from the same node.

Get the verification pack

Treated vs untreated activity curves, the node’s soil and air readings, and shareable charts for your sales team.

Request a verification pilot

What a node measures

Six readings on one timeline: microbial activity from the biology sensor, plus five soil and air readings, so activity can be read in context.

Soil temperatureAt the node

°C

Relative moistureCalibrated dry to wet; not a water-volume measurement

%

Carbon dioxide (CO₂)Concentration, not a flux estimate

ppm

SoilData.ai™ field node

Microbial activityΩ

The primary signal, from the printed biology sensor.

The five readings around it give context. They aren’t water-savings, irrigation-control, efficacy, or carbon claims.

Air temperatureAt the node

°C

HumidityAt the node

%RH

Same timelineAll six streams share one clock

1 node

The biology sensor

The biology strip is a printed polymer–carbon sensor. As soil microbes break down the material, its electrical resistance changes, which we use as a proxy for microbial activity and organic-matter decomposition at the node. It’s built for continuous field use: relative activity at one node, not a lab test of the whole field.

Treated vs untreated, side by side

Treated vs untreated nodes at one siteSix-week window · 4-hour readings

Microbial activity signal (relative)Treated nodesUntreated nodes

Reading the activity lineThe biology sensor reports electrical resistance (Ω). As soil microbes break down the sensor's material, its resistance changes, so we use it as a proxy for decomposition and microbial activity at that node. Moisture and temperature also affect the reading, which is why each node logs them alongside it. The line is drawn from that resistance signal and plotted so a higher line means more activity; it is not a standardized activity score. In a pilot, treated and untreated nodes at the same site are read the same way so they can be compared side by side. It isn't a yield, ROI, nutrient, or carbon number, and readings aren't meant to be compared across different sites. Values on this chart are illustrative.

Fair comparisons

Matched nodes at the same site, read with the same soil and air measurements.

A validation test at your site

We run the test at your site, measuring living-soil activity where the product was applied and on a matched untreated area. You get a site-level result you can put in front of a grower.

Science foundation

Built on research from Professor Gregory Whiting's Boulder Experimental Electronics and Manufacturing (BEEM) Laboratory at the University of Colorado Boulder (CU Boulder). Featured in NVIDIA's Earth2Studio soil-health blog (Aug 11, 2026), an early AI-modeling example. Not a live Earth-2 farm forecast, and not an NVIDIA endorsement or proof of product performance.