What FieldPoint does, and why it matters at write-up time.
Every feature below exists because of a specific way field data falls apart between collection and publication. Here is what FieldPoint does about each one.
A schema you design before you collect a single point
A soil nematode survey and a coastal erosion study have nothing in common except the app they're run in. Before the first point is collected, you build a field list for the project: choose from the auto-capture catalog, add your own custom fields, mark what's required, and set units. The next project starts from a blank schema too, not a leftover form built for someone else's fieldwork.
Mid-project changes do not rewrite history
Research plans change. If a supervisor asks for one more field in week six, FieldPoint snapshots a new schema version rather than mutating the existing one. Points collected under version one keep the fields they were collected with. Nothing already recorded silently changes shape, and you can always tell a reviewer exactly which version of the schema was active for any given sample.
The device already knows most of this. FieldPoint just writes it down.
These fields are captured automatically the moment a point is opened, with no extra taps required, and stored exactly as the sensor reported them.
Location
Latitude, longitude, and altitude in WGS84, plus horizontal accuracy, heading, and speed, useful for transect-style sampling and for reviewers who ask how precise a fix really was.
Barometric pressure
Read from the device barometer in hectopascals where available. Not every phone has one, so the schema builder flags it as possibly unavailable rather than failing silently.
Sun phase and position
Sun altitude, azimuth, phase, and local sunrise and sunset, computed on the device with no network call, so it works exactly the same with no signal at all.
Timestamps, both of them
UTC for reproducibility, local time for readability. Both are recorded on every point, not just whichever the phone happened to display.
Device and app version
Device model, OS, and the app build that collected the point, standard methodological disclosure that's easy to forget when it's not automatic.
Weather, opt-in per project
Temperature, humidity, pressure, wind, cloud cover, conditions, and UV index from OpenWeather, gathered in a batched sweep. See the dedicated section below.
Custom fields for whatever the sample actually needs
A device can report where you stood. It cannot tell you the species, the soil pH, or the tag number. Custom fields cover that half of the record.
Number
pH, temperature, count, concentration, anything measured with a unit.
Text
Species names, site descriptions, free-form notes.
Boolean
Present or absent, flowering or not, a simple yes or no observation.
Date
For anything that isn't the collection date itself, like an emergence date.
Select
A dropdown of options you define, so the same term is spelled the same way every time.
Photo
Attach one or more images to a point, saved locally with the filename convention below.
Offline-first, not offline-eventually
Points and photos write to the device the instant they're captured. There is no queue waiting for a connection and no risk of losing a sample because a dead zone swallowed a sync request. Weather is the one piece that needs the internet, and it is deliberately decoupled from collection itself, see the weather section below.
Provenance on every record, not an afterthought
Collector identity, device info, GPS accuracy, and both timestamps ride along on every single point automatically. It's the metadata a thesis committee or a journal reviewer asks for first, already attached rather than reconstructed from memory after the fact.
Collision-safe by default, even for a team of one today
Every record gets a globally unique identifier generated on the device, not a simple counting number. That single decision is what lets a second collector, or a second device, join a project later with zero migration.
Why not just count up from one
Counting IDs work fine for a solo researcher on one phone. The moment a labmate collects into the same project on their own device, both phones produce a "sample 14," and there's no clean way to merge them afterward. Unique IDs sidestep that permanently.
You still get clean sample numbers
A separate, human-facing sequential number rides alongside the unique ID, so your samples are still "sample 1, sample 2" for display and for filenames. Collector initials can disambiguate that numbering the moment more than one person is collecting.
Weather, gathered in a sweep, provenance disclosed
Weather is collected once per session in a batch, and nearby points collected close together in space and time share a single reading rather than each making an independent call. FieldPoint records whether a point's weather was queried directly or inherited, so nothing about the method is hidden from a reviewer.
Back up and restore without an account
Export a full backup and it goes straight to the OS share sheet, so it lands wherever you already keep files: your own Drive, iCloud, Dropbox, or a folder on your laptop. Restoring on a new device is picking that file and importing it. No login, no server call, and it costs nothing regardless of team size.
The .fpa archive is just a zip, on purpose
FieldPoint backups use a .fpa extension so tapping one opens it in the app, but it's a standard zip file underneath with a manifest that gets validated before anything is trusted: is the format right, is the version understood, does the checksum match. There is no encryption, deliberately, so a forgotten password can never lock you out of your own fieldwork.
Export that goes straight into your analysis
GeoJSON
A single FeatureCollection per project, with 3D coordinates so altitude renders natively in QGIS and ArcGIS, and every schema field included as a clean property.
CSV
One row per point, latitude and longitude as their own columns for a straightforward delimited text import, with ISO 8601 timestamps throughout.
Data dictionary
A generated field name, type, unit, and description table shipped alongside every export, ready to drop into a methods section or supplementary material.
See how it all fits together
The documentation covers the data model, the field catalog, and the weather methodology in detail.