Blankinship Soil Health Lab

Privacy policy

Last updated: 14 October 2026

This policy covers the Blankinship Lab Hub, the software of the Blankinship Soil Health Lab at the University of Arizona, at hub.blankinshiplab.com, its test copy at hub-test.blankinshiplab.com (section 7a), and this website, blankinshiplab.com. It says what the Hub keeps about people, why, and where. In it, "the Hub" means the software and "the lab" means the Blankinship Soil Health Lab.

1. What the Hub takes from your Google sign-in

You sign in to the Hub with a Google account. From that sign-in the Hub reads and keeps only:

Nothing else. Signing in gives the Hub no access to your own Google Drive, Gmail or Calendar, and the Hub never reads them.

2. Requests for access

When someone who is not yet an approved member of the lab signs in, the Hub records a request for access: the details in section 1, when the request was made, and how many times that person has signed in. The lab's Hub administrator approves or declines it. Declined requests are kept as a record; at present they are kept indefinitely.

3. Approved members

For each approved member the Hub keeps their name, e-mail address, role in the lab, home site, the role and rights they have in the Hub, and whether their account is active or disabled. These records are never deleted, so the lab's records stay attributed to the people who made them; a member who leaves is disabled, not removed. The Hub also keeps each Google sign-in linked to a member, with the details in section 1.

4. Sessions and cookies

When you sign in, the Hub stores only a one-way hash of your session's token, never the token itself, with when the session began, when it ends and when it was last used, and the kind of client you signed in with (a web browser, a phone or the desktop app) and that client's version. The Hub stores no IP address and no details of your browser.

An ordinary session ends after 12 hours, or after 60 minutes without use. If you choose to stay signed in, the session lasts 14 days. Sessions are never included in the Hub's own exports of its data (section 7 says what Railway's backups hold).

The Hub sets two cookies, both strictly necessary: one holds your session, and a short-lived one carries a sign-in while it is under way. The Hub uses no other storage in your browser.

5. The record of changes

The Hub keeps a record of every change made to the lab's records in it: what changed, from what to what, and who made it. The lab's work in the Hub, such as notes, attached files, photographed notebook pages, samples, shipments and datasheets, is attributed to the people who did it.

When a photograph is attached to a note, the Hub reads from it the time it was taken, the camera or phone model and, if the photograph records it, the place it was taken. It keeps these with the attachment, and so in the record of changes, as suggestions that become part of the lab's records only when a person confirms them. The photograph itself is kept in the lab's Google Drive as it was uploaded.

6. No analytics or tracking

The Hub uses no analytics, no trackers, and no scripts or fonts from other sites. This website likewise sets no cookies and loads nothing from other sites.

7. Backups

The lab keeps its own archive of the Hub's data, as files on the lab's computers; an administrator of the Hub may also download one to the computer they are using. It holds everything in the Hub's database except sessions, sign-in attempts, work in progress and the Hub's own keys.

The Hub's hosting provider, Railway, also makes daily backups of the Hub's database and keeps a continuous copy for restoring it to an earlier moment. These copy the whole database, including the session hashes, sign-in attempts and the Hub's stored authorization to the lab's Google Drive (which is kept encrypted), and are kept for Railway's retention period.

7a. The test copy

The lab also runs a test copy of the Hub at hub-test.blankinshiplab.com, where changes to the Hub are tried before they reach the Hub. Before each round of testing it is given a copy of the Hub's records as the lab's archive holds them (section 7): the people and their Google sign-ins, requests for access, the record of changes and the lab's work. On it everyone except the named testers is disabled, and the Drive ids of the lab's files are removed. It runs at Railway in the same US West region (California). Files it makes are stored in the Google Drive of a separate test Google account belonging to the lab, never in the lab's own Drive.

8. Transcription of attached images

A lab member may ask the Hub to transcribe the images in a file attached to a note: a photograph, the scanned pages of a PDF, or the pictures in a Word document. Those images are read from the lab's Google Drive and sent to the university's AI gateway, which returns the text; the draft and the accepted transcript are kept in the Hub's database. Nothing from a restricted project is sent to it. The gateway does not use what the Hub sends it to train any AI model.

9. Logs

The Hub's own log of requests records the path of each request, never the query that may follow it, and the address of Railway's edge server rather than yours. Railway keeps its own log of requests to the Hub and to this website, which records your IP address, your browser, the path and the timings, for 30 days. This website's own log records only the path of each request.

10. Files

Files uploaded to the Hub, and the files the Hub makes, are stored in the lab's own Google Drive, owned by the lab's Google account; on the test copy (section 7a) they are stored in the Drive of its test Google account instead. The Hub's database keeps their Drive ids and details, the text read from files attached to notes, transcripts of attached images (section 8), and page pictures of the lab's method files, so that they can be searched and shown. A workbook the Hub generates may be held in the database until Google Drive accepts it.

When a Drive file is linked as the source of one of the lab's methods, the Hub keeps that file's details from Drive, which may include its owner's name and e-mail address. That owner may be someone outside the lab.

11. How the Hub uses Google user data

The Hub works with Google Drive through one Google account belonging to the lab, and its test copy (section 7a) through one separate test Google account belonging to the lab; never through yours. Each has full access to its own account's Drive and uses it only to provide the Hub's features: the files and folders the Hub writes, reads and marks for the lab's work. The Hub never sells this data, transfers it to no one except as section 8 describes, and never uses it to develop, improve or train any AI model. No one reads it except in the course of the lab's own work.

The Hub's use and transfer of information received from Google APIs adheres to the Google API Services User Data Policy, including its Limited Use requirements.

12. Where the data is held

The Hub and its test copy run, and their databases are held, at Railway in its US West region (California). Where Railway keeps its backups and its continuous copy of the Hub's database (section 7) is not yet confirmed: the lab has asked Railway, and this section will say once Railway answers. The lab's own archives of the Hub's data are files on the lab's computers, or on the computer of an administrator who downloads one (section 7). The lab's files are held in the lab's Google Drive, and the test copy's in the Drive of its test Google account. Cloudflare answers for the domain blankinshiplab.com as a DNS service only: no traffic to the Hub or to this website passes through Cloudflare.

13. Questions and changes

Questions about this policy, or about what the Hub keeps about you: blankinshiplab@gmail.com.

When this policy changes, the new version is published here with its date.