Roadmap · HHT Diary

At a glance

Now, next and later.

Bugs, implementation progress and technical tasks live in GitHub Issues and Milestones; shipped versions are in the changelog.

Available today

Shipped and usable.

Personal diary: collection, inference, correctionsFoundation
Stats, widgets and survey-shaped exportFoundation
Calibration harness and regression fixturesFoundation

In progress now

Being worked on.

App Store release of the personal diaryFoundation · target around Nov 2026

Planned next

Intended, not built.

Granular, voluntary exportFoundation
Prompted recall and study configurationResearch & Analytics
Formal accuracy benchmark and export schemaResearch & Analytics

Exploratory later

Ideas, no commitment.

A research edition beside the personal appResearch & Analytics
Android, local languages and modesOpen Mobility Ecosystem
Self-hosted, offline-first study collectionOpen Mobility Ecosystem

Workstreams

Three tracks, one shared engine.

01 · Foundation

A personal diary people can rely on

Everything else depends on the core being trustworthy: recording that doesn't drain the battery, inference that is deterministic and documented, and corrections that are never overwritten. Getting the app onto the App Store matters because building from source keeps out almost everyone who isn't a developer.

Finer-grained export comes next because sharing your own data should never mean sharing all of it.

Background collection and deterministic reconstructionAvailable
Motion-aware GPS with a battery-saving stationary mode; visits, trips, mode segments and named places.
Corrections, stamps, stats and widgetsAvailable
Edits locked against rebuilds with an audit log, one-tap self-reported stops, period comparisons and Home Screen widgets.
Export and calibrationAvailable
CSV, JSON, GeoJSON and SQLite with a Python loader; a calibration harness and regression fixtures (#46); accessibility audits for the app and site.
App Store releaseIn progress
The same personal diary, installable without Xcode, with privacy disclosures that match the Privacy page. Target around 10 November 2026; it depends on App Review and may move. Version 1.0 adds no research features.
Granular, voluntary exportPlanned
Choose dates or trips and a level of detail (aggregates, trips without routes, full traces), make the SQLite snapshot follow the same choice, and preview exactly what will leave the phone. Today's behaviour is described in what an export contains.

02 · Research & Analytics

From a diary to a defensible dataset

Researchers need more than a trip list: they need answers to questions the GPS can't give, fixed study settings, consent, and evidence of how accurate the data is. Without those, TripLibre data is a curiosity, not something a thesis or an agency can rely on.

These features would live in a separate research edition so the personal app stays simple. None of them make TripLibre a representative survey; that still takes a study design.

Prompted recallPlanned
End-of-day questions about trips the app is unsure of: mode, purpose, companions.
Study configuration, consent and hand-overPlanned
A configuration file for questions, exported fields and study ID; built-in consent text; an explicit step that hands data to the researcher.
Researcher toolkitPlanned
Checks for incomplete days, pooling several participants' exports, aggregate-only summaries, and keeping the engine's original output when a record is first edited so accuracy can be measured from real use.
Validation and data standardsPlanned
A multi-participant accuracy benchmark with a published protocol, reproducible GPS fixtures, a battery and completeness benchmark, and a versioned export schema mapped to common HHTS codebooks. Today's measurements are on the Validation page.
Transit-aware mode detectionExploratory
Using transit data such as GTFS to tell bus from car.
How the research edition is packagedExploratory
A separate app, a separate build of this repository, or a study mode switched on by a configuration file.

03 · Open Mobility Ecosystem

Travel data within reach of any city

The places that most need travel data, often fast-growing cities with limited budgets and technical staff, are the least served by commercial tools. Reaching them means Android, local languages and modes, local calibration and partners who own the data. This is the furthest-out track, explained on the Vision page.

Any feature that sends data off the device would be opt-in, documented on the Privacy page, and discussed publicly before it is built.

AndroidExploratory
Most of the places this track aims at would need it first.
Local languages, modes and place typesExploratory
A multilingual interface with shared taxis, minibuses, motorcycle taxis and other local options.
Local calibration with partnersExploratory
Tuning inference on local travel conditions, with the calibration harness that exists today.
Offline-first hand-over and self-hosted collectionExploratory
For places with poor connectivity, and for studies that run their own server.

Two editions, one engine

A diary and a research instrument pull in different directions.

A diary should stay simple and ask nothing of its user; a study needs consent screens, fixed settings, questions and a controlled hand-over. Rather than pile study features onto the personal app, HHT Diary may split into two editions over time. How is not decided, and depends on App Store rules, recruitment and what partners need.

personal edition

Personal diary

Today's app and the App Store release. Automatic diary, quick corrections, stats, full export. Stays free and needs no account.

Available

research edition

Research support

For studies run by a researcher, class, agency or city: configuration, consent, prompted recall, selective hand-over, a researcher toolkit, and later Android and local languages.

Exploratory

shared core

Same engine and data

Both would use HHTCore, so a fix or calibration in one improves the other, and data from both can be analysed the same way.