Testing the performance of touch panels
Last week, the touch gestures and their importance for our user-experience were explained. Now, let's talk about which tests can be performed to warrant that touch gestures work properly on touch-enabled devices.
Since there are so many different tests, dividing them into different categories is a good idea. Let’s start with hardware-related performance tests.
Performance Tests
Performance tests are accomplished with robot-assisted platforms that mimic human interaction with touch-enabled systems. These platforms run on XY stages that use linear guides and encoders for the precise positioning of the Z-axis. The Z-axis is equipped with fingertips that accomplish the touch activity to perform the gestures and tests. The Z-axis may have more than one finger, and in that case, the system includes an azimuth unit such as shown in the below example.
Some examples of performance tests
Tap Tests
These tests measure the DUT's (device under test) tap accuracy and repeatability performances. The results are reported with the maximum accuracy and repeatability errors and missing inputs values. The results are used to determine the hardware’s ability to report consistent coordinates, especially when the touch panel's same location is pressed repeatedly. Without going into more details, here are some sample test results for one finger tap accuracy and repeatability tests:
The same accuracy and repeatability tests can also be performed with multiple fingers. Here is an example of that:
The same accuracy and repeatability tests can also be performed with multiple fingers. Here is an example of that:
As a bonus:
From here, you can read a case study in which human touch’s accuracy and repeatability are tested.
Swipe Test
The test measures the DUT's accuracy when swiping in a linear motion. It reports the maximum allowed values of offset, jitter, and missing swipes.
Without going into more details, here is an example analysis for multiple fingers swipe test:
Jitter Tests
These tests measure the DUT's performance when a single point is pressed and hold. It reports the maximum measured stationary jitter calculated from the first coordinate reported by the DUT.
Here how a failed jitter test’s result looks like
Reporting Rate Tests
These tests measure and report the DUT's stationary and non-stationary reporting rates.
Here how a stationary reporting rate for an individual tap looks like:
In this result, the X-axis is the DUT’s event index, and the Y-axis shows the delay from the previous event. It is also common to test the non-stationary reporting rate calculated from time intervals between reported coordinates from DUT. The amount of missing lines (missing inputs) is also interesting to analyze as part of these tests.
Various other tests can be considered as performance tests on the panel level. Such tests are for analyzing the First Contact Latency and Hover, for example. As part of the first contact latency test, the maximum allowed response latency from active and idle states are measured. As part of the hover test, jitter, missing swipes, and the maximum permitted hover distance from the DUT’s surface are measured.
There are some standard tests performed just with multi-finger setups. One of them is the separation test that reports maximum allowed finger separation distances in vertical and diagonal alignments.
Summary
Smooth user experience relies on accurate and repeatable touch gestures, which are only possible when the touch user interface (UI) is functional, responsive, and fast over the entire surface area. The manufacturers of touch-enabled devices need to perform various tests on the system, device, and individual component level to fulfill the end user’s expectations. In this article, some of those tests are mentioned.
Next week, I will write about which tests are used to test the user interface's performance on touch-enabled devices.
With Regards,
All images are courtesy of OptoFidelity.