Red Pebble | John Ozbay

Red Pebble, a Google Assistant & Google Home bot, that tells you everything about the weather on Mars!


How does Red Pebble work?

Red Pebble uses data from Centro de Astrobiologia (CSIC-INTA) using the amazing API kindly provided by {MAAS} API

The weather & atmospheric data is being transmitted by the Curiosity Rover on Mars. Red Pebble then caches this atmospheric weather information daily using Google Cloud Functions on its own servers to make sure that our assistant doesn’t make tons of requests to the kind providers of the API listed above.


The cached version of the API can be found here :

Red Pebble is built using API.AI, Google Cloud Functions and Firebase. It has two types of answers:


Red Pebble

+ Informative interactions that doesn’t require the use of the weather API. These are questions like : “Are there any clouds on Mars?” or “Does it rain on Mars?” etc. These types of questions, that doesn’t require any up-to date weather information are handled by


+ Interactions that require the use of the weather API. These are questions like : “What’s the temperature on Mars?” or “What’s the season on Mars?” etc. These types of questions are first received by, which in return makes a call to cloud functions server that fetches the latest info from red pebble’s cached latest.json file and finally return the answer to the user.


The Cloud Function also logs anonymous user IDs to change the way it responds to save them time and frustration. This part is handled by Firebase.


Where did the idea come from?

One day Shelby Hutchison, Senem Cinar and I were having breakfast and we thought wouldn’t it be hilarious and awesome to have a home assistant that tells the weather on Mars. Then for 7 days after work, we all went home, didn’t sleep, designed and coded an assistant that does just that.



Centro de Astrobiologia (CSIC-INTA)
And everyone from the Google Home Hackathon NYC 2017 for helping and supporting us.


Open Source

Red Pebble is 100% Open Source, with MIT License, available on Github.