The ability to accurately and consistently measure human attention span may be an important factor in early detection of attention deficit disorders, Alzheimer’s disease and other neurological health assessments. Present measurement techniques rely on direct observation and testing the subject’s ability to focus or provide information on how easily the subject can be distracted. Experts have noted that different methods and variability in the testing environment can produce quite varied results, often resulting in misdiagnosis and overmedication.

What is needed is a method to assess human attention span in manner that is quantitative and repeatable and equally applicable to children and adults.

The Cube-it project

What is proposed is system that includes a tabletop base unit and a wirelessly connected interactive cube that combined, presents the subject with a series of timed stimulus and response tasks. Audible challenges like "Press the red side" or "Tap the blue side twice" would initiate a task, with the system recording the time it takes to complete the task and how the subject is handling the cube.

The goal of the system would be to create a data set of measurements that could be compared both to absolute scales (percentile of population, etc.) and relative to prior testing (this subject is showing greater/lesser attention span).

Who do we need to be successful

  • Some coders that have experience with Bluetooth connectivity between an Arduino or Teensy and a Raspberry Pi
  • Hands-on builders that can solder, saw, swear, and generally make things fit
  • Someone that is or knows a fidgety person to help get the initial test sequences reasonable
  • Someone with a soft, soothing, melodic voice to record the verbal prompts. Otherwise, someone who knows how to get a synthetic speech engine going that has a soft, soothing, melodic voice built-in.
  • Someone that can help think through the command structure from the base to the cube
  • Some folks with experience with Arduino code, maybe some Python or similar for the Raspberry Pi
  • Someone with experience decoding and recording accelerometer data

What's going on inside

  • Bluetooth connection between cube micro and base
  • Individually programmable LEDs to light up each side of the cube
  • Accelerometers to measure cube orientation and movement
  • Headphone connection on base
  • Raspberry Pi (or similar) in the base creating the test sequences and audio cues
  • Wifi connection from base to smartphone or computer to display results
  • Batteries, switches, power supplies, wires, solder, coffee residue
Share this project:


posted an update

A big thank you to the IoTHackDay 2017 crew for setting up this year's event! We got the Raspberry Pi 3 running a basic Python program to read the streaming accelerometer data from the cube, process switch events and send out button-light-up commands. Bluetooth was a bit of a pain, so that needs to be made easier. Next steps: a) Inductive charging for the cube b) minor tweaks to the communication protocol between the devices c) improve the robustness of the Bluetooth connection and d) increase the sophistication of the Python command and storage program.

Log in or sign up for Devpost to join the conversation.

posted an update

Just a quick update (T-minus 4 days!!!) - we've accumulated quite a few hardware bits and pieces (Bluetooth devices, Raspberry Pi boards, batteries, switches, LEDs, wire, etc) and I've taken a stab at 3-D printing some cube parts so we can focus on final assembly and of course, the programming. That's where we can use help. If you've played with Arduino code, Bluetooth connectivity to a Raspberry Pi, building databases of real-time streaming data from a fidget cube, (just for example), please sign-up or simply swing by. See you there!!

Log in or sign up for Devpost to join the conversation.