Weightless

Hardware Job Proposal

Goals

  • Make a prototype
  • very interested in lean: short timeline to prototype and low initial development cost
  • Wiling to sacrifice on per unit costs
  • size: small, quarter of a hockey puck is ideal 
  • battery life: long, a year long is ideal, but aware that might not be possible. This early on, I'm okay with a month battery life for the prototype

Description (nontechnical)

I want to make a device that attaches to weights in a gym, and creates a seamless workout tracking experience for people who attend the gym. All the end user needs to do is walk into the gym, turn on the app and Bluetooth and workout. That's it. No pairing or anything. The user should be able to see how the workout is going in close to real time if he/she wants to, tracking repetitions, movements, form, etc. The user would also be able to access that information after the workout. Think RunKeeper or MapMyRun, but for the gym.

How might this work?

I'm thinking a solution with the following features

  • iBeacon (via Bluetooth 4.0)
  • Wi-Fi
  • 3 axis- Accelerometers
  • 3 axis- GyroScope
  • 3 axis- Magnetometer, maybe?
  • Battery operated
  • Small dimensions, the smaller the better
  • Willing to sacrifice on battery life
  • Replacable battery probably, but willing to explore other options
    • also considering just battery, not even replacable
    • replacable might be better though for initial iterations

iBeacon will detect who is close by in terms of device ID. The Wi-Fi module will transmit the raw data to my server, where the real number crunch will happen (analysis of the workout- reptitions, movement, form). The accelerometers, gyroscope, and magnetometers will record the movement data.

With all of this, we can achieve the seamless workout.

Thoughts on Implementation

Here's what I have come up with. But willing to discuss other strategies :)

  • Microcontroller -> Particle Platform
    • Set of chips that have built in Wi-Fi and the ability to deploy new code over Wi-Fi
    • Spark Core - ships immediately $40 per
    • Photon - $19 per, ships in June (has a lead time)
    • P0 and P1 - $10 / 12 per or less based off of quantity (6-8 week lead time)
    • specs are here https://www.particle.io/prototype
    • ARM-based
    • could save development time & money
  • No opinions on regarding the BLE, accelerometers, gyroscopes, and magnetometers
    • although I do appreciate the IMUduino and what's on there http://femto.io/products/imuduino
    • not thinking of putting Spark and IMUduino together, just posted IMUduino for parts
    • really would like great accuracy, have no idea whether cheaper components would do fine, or these are it, or better can be done 
    • FreeIMU, the components and the library also look cool, really cool http://www.varesano.net/projects/hardware/FreeIMU

Would like to know the following:

  • Prototype Implementation (very interested in lean: short timeline to prototype and low initial development cos, Wiling to sacrifice on per unit costs)
    • Number of weeks for development of prototype
    • Cost of development of prototype
    • Per unit cost of prototype if scale
    • Dimensions of prototype
  • Your ideal implementation (not consicostering of development and timeline -> smallest per unit cost, size)
    • Number of weeks for development of implementation
    • Cost of development of implementation
    • Per unit cost of implementation if scaled
    • Dimensions of implementation

You can contact me at:
RRP54@cornell.edu
201-316-4184.