Data Sources

Every source used to power Oahu commute predictions

Community Commute Data

Core source

Real Oahu commuters

The most important data source in AlohaShift. Real commute times submitted by Oahu drivers are the ground truth that no API can replicate. Every report contributes to more accurate predictions — not just for that exact route, but for all routes sharing the same road corridors.

How one report helps many routes

When you submit a commute from Pearl City to Downtown Honolulu, we extract which highways your route used — H1, Moanalua Freeway, etc. That data then improves predictions for any other route using those corridors at the same time of day: Ewa Beach to UH Manoa, Aiea to Kapiolani Medical Center, and more.

Step 1You submit departure time, arrival time, and your route

Step 2We calculate actual travel time (arrival − departure)

Step 3Google Directions API identifies which Oahu corridors your route uses

Step 4Report is saved to database with corridor tags (H1, Pali Hwy, etc.)

Step 5Future searches on overlapping routes draw from your data to calibrate predictions

Corridors tracked

H1, H2, H3, Pali Highway, Likelike Highway, Kalanianaole Highway, Kamehameha Highway, Nimitz Highway, Farrington Highway, and more

Submit your commute data

Hawaii DOE School Calendar

Seasonal context

School day detection

Oahu traffic during school breaks is dramatically lighter than on school days. The same Monday at 7 AM can differ by 20–30 minutes depending on whether school is in session.

Every community commute report is automatically tagged with whether school was in session that day. School days and non-school days are stored separately, so future predictions always compare like with like.

SourceHawaii DOE Academic Calendar (SY 2024–25 and 2025–26)

Breaks coveredSummer, Winter, Spring, Thanksgiving, Federal & State holidays

Applied toEach commute report at submission time

UI indicatorSchool day / School not in session — shown in search results

Hawaii DOE School Year Calendar

Google Maps Distance Matrix API

Primary

Travel time baseline

For each departure time slot, we query Google Maps with traffic_model=pessimistic to obtain predicted travel durations.

Endpointmaps.googleapis.com/maps/api/distancematrix

Traffic modelpessimistic — worst-case historical scenario

Also used forDirections API — extracting route corridor names from commute reports

Update frequencyReal-time query per search

Known limitation: Google Maps underestimates Honolulu peak-hour travel times. Real commute data showed 62 min actual vs 30 min predicted — a 2× gap.

Google Maps Distance Matrix API Documentation

HDOT HPMS Dataset

Government open data

Commuter count baseline

HDOT publishes Annual Average Daily Traffic (AADT) counts for all major state highways. We use this to calculate the realistic number of Oahu morning commuters.

DatasetHPMS — Highway Performance Monitoring System

APIhighways.hidot.hawaii.gov/resource/3jb9-z582.json

Key findingH1 peak section AADT = 65,800 vehicles/day

Derived estimate~10,857 Oahu morning commuters (6–9 AM)

Calculation: 65,800 vehicles/day × 15% (morning rush share) × 1.1 (avg occupancy) = 10,857 commuters

HDOT Open Data Portal

TomTom Traffic Flow API

Supplemental

Real-time speed validation

We use TomTom's Traffic Flow API to cross-check real-time Oahu road speeds against free-flow baselines.

Endpointapi.tomtom.com/traffic/services/4/flowSegmentData

Update frequencyEvery 2 minutes (real-time)

TomTom Traffic API Documentation

AlohaShift is an independent student project built for the 2026 Congressional App Challenge. Not affiliated with Google, TomTom, or HDOT. All data is used in accordance with respective terms of service.