IoT Enabled Smarter Home

Software Defined Speaker, Alarm, Watering, Security Camera... = Software Defined Everything!

Next generation smarter home is coming - and I'm building a super IoT connected intelligent personal assistant at home for busy professionals, old people, chronic patients etc.

Demonstrable for HK Speakers:

  1. Internet Radio - any http based Internet streaming radio

  2. real-time environment Sensors (temperature, humidity, noise / ambient sound, barometer, ambient light, any you can have more, e.g. smoke, gas, air quality...)

  3. TFL real time bus API -> Text to Speech -> playback

  4. Beacon geofencing - e.g. when you are about to leave the door, it will tell you the latest bus or train info!

How it works

  1. Intel Edison Wifi --> internet
  2. Intel Edison Bluetooth --> Harman Kardon speakers
  3. Intel Edison Bluetooth Low Energy --> can talk to BLE/(i)Beacon Stickers + Wearable sensors
  4. TI SensorTag as a dev watch device, including 9-axis movement sensors!

leverage Node.js eco-systems

Intel Edison Node.js engine --> API calls

  • Transport for London (TFL) real-time bus API
  • Google Translate Text to Speech API
  • PulseAudio + Lame for mpeg decoding and audio playback

Challenges I ran into

  1. Google only returns 8192Hz audio - took 1 hour midnight to figure out!
  2. TFL API sucks :(
  3. Multiple Asynchronous HTTP API calls

Accomplishments that I'm proud of

  1. My Harman Kardon Speaker can actually speak, about anything!!
  2. I have first laser cut front panel for my SmarterHome last Friday!

What I learned

  1. more Javascript tricks - with Node.js running on embedded Linux platform
  2. more APIs for Smarter Home - this time Google Translate API, TFL Bus API

What's next for Smarter Home (with Software Defined Speakers)

  1. I'm Laser Cutting a proper case - so I can make a beautiful prototype before going to KickStarter to raise money to productions the solution!
  2. TI SensorTag IMU - 9-axis sensor is working with Intel Edison BLE, collecting data for gesture detection - so that we can use Gesture to control HK speakers!
  3. Research into Internet Audio - so hopefully HK can embedded that functions into the speakers to make it even more hackable!
  4. Following up with HK / Kevin's team to explore upcoming non-mobile APIs to further integrate speakers with Smarter Home.
  5. Research into Q&A AI systems for home care use cases

Built With

Share this project: