Human learning abilities in using different types of physical human-computer interfaces

26th Marian Smoluchowski Symposium on Statistical Physics Complexity of Brain – Critical Behavior and Scaling Kraków, Poland, August 28–31, 2013

Michał Sznajder – Jagiellonian University, Krakow, Poland dr hab. inż. Jakub Barbasz – PAN, Jagiellonian University, Krakow, Poland

Background Aim of the study Methodology Planned results

Human-computer interfaces are the key elements in our everyday use of modern consumer electronics and computers.

Computers are becoming more and more fluid - they move with us, they travel with us in shape of mobile devices.

On top of this innovative physical interfaces using inertial sensors, change the way of using and thinking about electronic devices and make them more natural and direct by using movement.

Development of this branch of interface design is a follow up of popularization of physical sensors (accelerometer, gyroscope and magnetometer) in consumer electronics.

In case of modern consumer electronics (smartphones, tablets, game controllers, etc.) building effective human computer interface is a key factor of product market success.

Design problems:

What parameters should designers control?

How to choose interface parameters values so it will be effective.

How to make it easy to use and easy to learn?

What factors decide about interface being good (good = does not create second problem)?

The purpose of the study is better understanding of how human adjusts and learns to use different kinds of inertial human-computer interfaces.

In other words: how human learns to use different types of movements in solving given problem and which kinds of movements are easier to incorporate in this aim.

In the study three different physical human-computer interfaces were designed and built using three most popular inertial sensors:

Accelerometer Gyroscope Magnetometer

All this was built using consumer electronics sensors incorporated in a smartphone.

Gyro mouse

Allowing user to control the device with tilting a hand.

Controlled device reacts on every angle change and pauses in movement.

Accelerometer tilt mouse

In this kind of interface controller device also reacts on hand tilt but using changes in Earth gravity direction not reacting directly to changes in tilt angle

Not reacting on pauses in hand movement.

Sensor fusion mouse

Sensor fusion mechanism incorporating signals from all the three sensors.

Computer mouse features + rotation

As a study reference point touchpad interface was also created.

All interfaces are implemented and tested on the same device.

Example simple movements

We can clearly see noise in the signal

Filtration level of sensors signal is one of the parameters interface designers have to control

Problem: we want to filter redundant information and keep movement information

Other problems solution: sensor fusion

+ +

Thanks to created interfaces we have full control of creating different flavors of inertial interfaces and full control over them – a situation similar to designer creating product interface.

Two way testing:

Whole test group will use special test arcade game measuring participants reactions time using different kinds of interfaces and speed of learning process (controlled device).

At the same time we will make measurements of movements precision when making specified kind of task over time (controller device).

Tests are in progress

Time is needed to reach statistical validity

Conducted research will help to choose proper solutions for specified kind of devices and targeted groups of users.

Especially it will help to point human computer interfaces key elements which decide about usability and learning speed.

This will help to create important tool for decision making on interaction design level

Applications:

Mobile devices

Game controllers

Augmented reality technology

Medical aid controllers

[email protected] michalsznajder.com