EW Project 2: Presentation and Conclusion

We brought all the kit in to do the presentation with and did a decent amount of testing beforehand and all seemed to work pretty well.

Here's a link to the presentation: http://goo.gl/nIrJAz

The demo went okay. I feel as though we didn't communicate well enough the wireless aspect of it. The two Arduinos were messaging each other using MQTT over WiFi and the internet. However as both were plugged into my laptop for power, it wasn't really something that we felt was demonstrates effectively. There was also one hiccup with a portion of the demo, which was when I hit the keyboard I dislodged the sensor from the bottom of the keyboard. Meaning when Jamie went to do his portion of demo - to demonstrate profiles - it didn't detect the hit. It all worked fine when we put it back in place though. Another issue was that because we used the speakers that DAT had, they didn't work immediately when powered they went from off straight to standby, so the music didn't play when Jamie hit the keyboard. Besides that I felt that we conveyed the concept fairly effectively.

Some more of the ideas for the future include:

  • Being a product designed to work with nearly everything, I feel that being more integrated with digital APIs and standardized physical sensors/buttons/switches would be essential for future development. So once activated it could connect to online services and do something broader than turning something on or off. Another one for the physical world could be a servo that's attached to a dial on an oven, or some speakers that changes once activated. 
  • Multiple forms of activation is also something that would be essential across different profiles. We used the keyboard hit for the demonstration, however other means of activation could include 'not typing for a while', or using facial recognition to profile the user's moods activation using different means. 
  • Multiple forms of profiling would be good too. For the demo we had to use a light sensor to detect the difference in opacity between the mug and the glass, but ideally we would have liked to use RFID. With each drink's receptacle having a different unique ID. We also discussed different forms of identification using things like weight sensors under the chair. 
  • A minor one, but different profiles being detected and making different type of coffee depending on the user's preferences.

I'm really happy with how this project went. Learning MQTT was so handy, and Chris' wifly sketch made using the wifly shield really easy. These two were probably the main things I took away from this project from a technical standpoint. using MQTT proved so much more stable and reliable than the method we used in the ELDR project where the Arduino just sent an HTTP request to a webpage.


Popular posts from this blog

Logitech G502 Proteus Core Teardown & Scroll Wheel Replacement

Turning a laptop screen into an external monitor.

Letter basket: Productive procrastination