Field notes

Short records from real evaluations: what we tested, the conditions we tested it in, and what we looked for. Detailed results, clients, and engagement details are not published.

Networking

Mesh networking in a dense urban environment

How well does a mesh network hold up among buildings and at street level?

We tested a mesh network among buildings and at street level, recording where links held, where they dropped, and how the network recovered.

The test used Meshtastic, a widely available open-source platform that sends short text messages over low-power LoRa radios, with no cell service or internet needed. We framed it around an ordinary use: keeping a team of event volunteers in touch across a busy downtown area, where phone networks often congest. Handheld nodes paired to phones over Bluetooth, and a few fixed nodes placed at height relayed messages between blocks.

Detailed results from this test are not published.

What we looked atHow message delivery held up as nodes moved between buildings and street level, and how the mesh rerouted traffic when a relay dropped out.
Sensors

Environmental sensing to inform field decisions

How far can atmospheric and environmental sensor data be trusted as a basis for decisions in the field?

We tested sensors that gather atmospheric and other environmental data, and evaluated how far that data can be trusted as a basis for decision-making in the field.

The test used widely available sensors that log temperature, humidity, air pressure, wind and air quality, the kind of readings that inform ordinary outdoor planning. We framed it around an everyday use: deciding whether conditions at an outdoor site are safe and workable for a crew, and when to adjust a schedule as the weather turns.

Detailed results from this test are not published.

What we looked atHow closely sensor readings tracked actual conditions on site, and how much confidence they could reasonably carry when a decision depended on them.
Engagement

Education and field evaluation in one engagement

What does it look like when instruction and evaluation are run together rather than in sequence?

An engagement paired practical instruction with hands-on evaluation of several emerging technologies. The participants learned to operate the devices; in the course of learning, they also generated a grounded read on usability and limits that a bench test would not have produced.

Running the two together has a specific advantage. A device's real weaknesses tend to surface not in expert hands but in the hands of a competent newcomer under time pressure — which is also the population most device specifications quietly assume away. Defaults that an expert silently corrects, steps that are easy to get wrong, indicators that are read the wrong way: these appear immediately when ordinary users are learning, and they are exactly the failures that matter in use.

The output was twofold — trained users, and an evaluation of the technologies that reflected how they behave for people who are not their designers. We regard the second as inseparable from the first.

What we recordedUsability is a performance characteristic, not a nicety. The engagement produced a plain account of where each technology helps a capable user and where it fights them.

These notes are illustrative of the kind of work we do and the form our findings take. Specific engagements, participants, and clients are not published. For a fuller account of how we reach a finding, see Method & standards.

Editor's noteThese notes were developed with the assistance of AI writing tools, then reviewed, revised, and approved by Technology Research Group.