Revolutionizing Analytics: AI and Natural Language to Query 1 Billion Flight Records

Try HeavyIQ Conversational Analytics on 400 million tweets

Download HEAVY.AI Free, a full-featured version available for use at no cost.

Big geodata has historically been a big pain to access and use. HeavyIQ changes that by providing what we call “Conversational Analytics”. This lets you query data with natural language. We convert your questions to SQL, run that on our GPU-accelerated database, and return answers in English, tabular and graphical form. Unlike some other platforms, you have complete control and traceability. We only use the data you have selected, and we show you exactly how it is being processed to answer your questions. Coincidently, this is also a great way to teach yourself SQL or enhance your existing skills.

‍

‍

HeavyIQ debuted back in February and we initially exposed 16 months of historical X/Twitter data. Since then, we have seen thousands of questions submitted and answered. When we found out that the renowned Mark Litwintschik was going to be featuring HeavyIQ and the HEAVY.AI platform in his blog, we decided to capitalize on this opportunity and introduce broader access options and updates.

‍

So today, we are happy to announce the addition of 35+ years of airports, aircraft and individual flights data which Mark loaded for his blog. This is about 220 million records of analysis-ready data sourced from United States Bureau of Transportation Statistics (BTS) and Federal Aviation Administration (FAA). As a bonus, we are also exposing over 1 billion Automatic Dependent Surveillance - Broadcast (ADS-B) records.

‍

Just in time for summer travel season, you can ask any question you might like about historic airline performance from your local airport or any other. To get you started, we’ve provided some examples below.

‍

You have three options to access this information:

Use the demo links and sample prompts below to access environments where the data is ready to go.
Create a local install, including the analysis-ready data. This lets you customize data and dashboards, but with the data wrangling already completed.
Download HEAVY.AI and follow the steps in Mark’s blog. This teaches you how to generate analysis-ready data from CSVs. This is applicable not just for HEAVY.AI, but also to most any database.

‍

Option #1: Try Live Online

‍

BTS & FAA Data: https://demo-flights-heavyiq.heavy.ai/flights/sql-notebook
- Show the percentage of flights canceled for weather related reasons by month of year and hour of day.
- For each airline with over 300,000 flights in 2023, calculate the average air time per day (365 days in a year) per aircraft for each year since 2014. Consider only flights that weren't canceled.
- Show distinct routes flown by Virgin America in 2016 which were NOT flown by Alaska Airlines in 2023.

‍

ADS-B Data: https://demo-adsb-heavyiq.heavy.ai/adsb/sql-notebook
- Show a histogram of number of records by altitude (bins of 1,000 feet between 0 and 60,000 feet) and speed (bins of 10 knots between 0 and 700 knots)
- Show the location of all Ryanair flights over Ireland, United Kingdom, France, and Spain on February 22, 2022.
- Beginning with the year 1930 and ending with 2020, return the average speed of aircraft by their decade of manufacture, in order of decade.

From here, you can access both of these flight datasets or the tweets which we introduced back in February.

‍

Option #2: Run Locally

‍

If you want to combine this data with your own, adjust dashboards, or otherwise have full control, you’ll want a local install. Fortunately, that’s exactly what our “Free” product is designed to support. As you’d expect, there are 2 steps - downloading/installing the software & restoring the data. Some basic docker and linux debugging skills will make this process easier.

‍

Download HEAVY.AI here.
Once you have installed the software, run the following commands from the SQL Editor to load the needed tables and data
1. RESTORE TABLE aircraft FROM 's3://batteries-included/flights_demo/dumps/aircraft.dump.lz4' WITH (compression = 'lz4', s3_region = 'us-east-1');‍
2. RESTORE TABLE airports FROM 's3://batteries-included/flights_demo/dumps/airports.dump.lz4' WITH (compression = 'lz4', s3_region = 'us-east-1');‍
3. RESTORE TABLE flights FROM 's3://batteries-included/flights_demo/dumps/flights.dump.lz4' WITH (compression = 'lz4', s3_region = 'us-east-1');‍
4. RESTORE TABLE usa_states FROM 's3://batteries-included/flights_demo/dumps/usa_states.dump.lz4' WITH (compression = 'lz4', s3_region = 'us-east-1');

If you need help with any of this process, access the community.

‍

For #3:

‍

Follow Mark’s blog. If you have any issues while walking through the process, access the HEAVY.AI community.

‍

---

‍

Here are a couple of examples of the types of visualizations you can automatically generate with HeavyIQ and the flight data. HeavyIQ doesn’t just do text-to-SQL, it also automatically generates applicable visualizations.

‍

With the BTS & FAA data . . .

Show the percentage change in flight arrivals per US state from 2019 to 2023.

‍

With ADS-B data . . .

Show the location of all planes where the ADSB country is NATO

‍

Dr. Michael Flaxman

Dr. Michael Flaxman is HEAVY.AI's Spatial Data Science Practice Lead. At HEAVY.AI, Dr. Flaxman’s team is focussed on the combination of geographic analysis with machine learning, or “geoML.” He has served on the faculties of MIT, Harvard and the University of Oregon. Dr. Flaxman has participated in GIS projects in 17 countries. He has been a Fulbright fellow, and served as an advisor to the Interamerican Development Bank, the World Bank and the National Science Foundation. Dr. Flaxman previously served as industry manager for Architecture, Engineering and Construction at ESRI, the world’s largest developer of GIS technology. Dr. Flaxman received his doctorate in design from Harvard University in 2001 and holds a master’s in Community and Regional Planning from the University of Oregon and a bachelor’s in biology from Reed College.