Sensors & power-users

i213 nick merrill fall 2014

ffff@berkeley.edu

sensors 

1:20

Sensors applications you know or use?

https://jawbone.com/blog/napa-earthquake-effect-on-sleep/

ecg (heart)

LifeBEAT

BASIS

gsr (skin conductivity)

emg (muscle)

eeg ('brainwaves')

eeg ('brainwave')

eeg ('brainwave')

are you thinking x or y? accuracy compared to computational efficiency

sociometer

Choudhury, Tanzeem, and Alex Pentland. "Sensing and modeling human networks using the sociometer." 2012 16th International Symposium on Wearable Computers. 

n=1 sensing: jawbone, fitbit, ..., maybe some sharing features but egocentric application goals

- heterogeneous sensors

- naturalistic (uncontrolled) conditions

- 'swarm' computing (lots of failure prone nodes)

sensing-at-scale 

better understanding of humans?

better predictive models of behavior/response?

methods that supplement/supplant lab-based ?

Social sensing: How can sensors help us transmit our lived experiences to other people?

prototyping with sensors

• bitalino: http://bitalino.com/
• adafruit generally: https://www.adafruit.com/
• conductive fabrics: https://www.sparkfun.com/products/10056
• eeg: neurosky mindwave, interaxon muse
• http://github.com/indra-net/
• femtoduino: http://www.femtoduino.com/
• building wireless sensor networks: http://www.amazon.com/Building-Wireless-Sensor-Networks-Processing/dp/0596807732

Part 2: power users

Power users will spend a lot of their lives in your software.

Bargain: With some user training, we can trade intuitiveness for efficiency.

Model 1: It's impossible to use the interface without training.

pros/cons?

Model 2: An "invisible layer" for power-users (Gmail, Google Calendar, Feedly, VLC, MacOS, [..])

pros/cons?

1. Advanced features are discoverable by accident.
    
2. Advanced features bootstrap on relevant knowledge & models.
    
3. Advanced workflows are similar enough to normal ones that non-power-users can follow along.

Guidelines for the "invisible layer"

4:55