December 16, 2016

Evaluation

Fall Validation Experiment:

Test 1: Position Control

The system performed as specified. It held the user’s weight (~80kg), and did not throw the user off-balance for a variety of different postures and positions of the user’s centre of mass.

Test 2: Velocity Control

While the skates were controlled to a constant velocity and the user could walk on them, the control was imperfect and tended to a slightly open-loop behaviour which was rectified before the encore.

Fall Validation Experiment (FVE) Encore:

Test 1: Position Control

Again, for the position test, the system performed as specified and held the user’s weight, did not throw the user off-balance for a variety of poses including squats, leaning in the fore-aft and medio-lateral directions.

Test 2: Velocity Control

Improvements were made as per feedback received at the FVE, and the system worked as desired. Success conditions were achieved; the user did not fall.

Spring Validation Experiment:

Success conditions were achieved for each case.

Test 1: Number of steps taken

On comparison between the number of steps taken with and without the system, we find that with the system, the user can take at least 6 times the number of steps taken without the system.

Steps taken without the system = 6 steps

Steps taken with the system = 88+

Test 2: User’s Final position

The system accurately and reliably recenters to within 75 cms of the initial position

Test 3: User comfort

The user does not feel any major discomfort due to the VR display; there are no quantifiable metrics for user comfort outside of the user’s word. Koushik’s recorded response was that the system is comfortable to use for intervals shorter than 10-15 minutes.

Test 4: User balance on the system

The system demonstrated 3 steps, bending and leaning sideways to display no misclassifications and no jitter in the system.

Spring Validation Experiment (SVE) Encore:

Test 1: Number of steps taken

On comparison between the number of steps taken with and without the system, we find that with the system, the user can take at least 6 times the number of steps taken without the system.

Steps without system = 6 steps

Steps taken with the system = 49+

Test 2: User’s Final position

The system accurately and reliably recenters to within 75 cms of the initial position,  which is better than the 1 m metric set by the team as a mandatory requirement. 

Test 3: User comfort

There are no quantifiable metrics for user comfort outside of the user’s word. Koushik’s recorded response was that the system is comfortable to use for intervals shorter than 10-15 minutes.

The user does not feel any major discomfort due to the VR display which was verified by unbiased users (Other MRSD graduates).

Test 4: User balance on the system

The system demonstrated standing, stepping, bending and leaning sideways to display no misclassifications and no jitter in the system.