Use different motion tracking technology

[osrf-migration]

Original report (archived issue) by David Kluger (Bitbucket: DKluger).

---

Infrared motion tracking, frankly, is bad for this application. The occlusion of markers is a huge problem for tracking the wrist, especially when the wrist is pronated/supinated. This problem is exacerbated when we must also accommodate for left-handed users (if/when this functionality is added), where it would be 100% necessary to change the marker positions on the head as well as move the camera to the left side of the user to capture the wrist. Magnetic motion tracking, for one, eliminates these problems.

[osrf-migration]

Original comment by Nate Koenig (Bitbucket: Nathan Koenig).

---

We have experience similar problems while testing simulation with Optitrack. Fortunately, we started our development using a Polhemus, and it is still possible to use polhemus with Gazebo. Do you by chance have a Polhemus?

We'll work on a tutorial that describes how to use Polhemus with Gazebo.

[osrf-migration]

Original comment by David Kluger (Bitbucket: DKluger).

---

The only mocap option we have is the IR mocap through the OptiTrack provided to us. Please let me know through this thread if use of Polhemus provides significant benefits.

[osrf-migration]

Original comment by David Kluger (Bitbucket: DKluger).

---

Through the grapevine, I have heard that DARPA opted to use IR mocap over magnetic mocap for this project for reasons that escape me. Can you inform me as to why this decision was made?

[osrf-migration]

Original comment by Nate Koenig (Bitbucket: Nathan Koenig).

---

We have found Polhemus to be quite effective. Data from the Polhemus has low latency, and is just as accurate (if not more so) than the OpiTrack. We are new users of magnetic tracking, but were very pleased with the results.

Here are some of the reasons that a magnetic tracking solution was not used:

1. Polhemus is much more expensive than OptiTrack.
2. Each magnetic "marker" is tethered by a cable.
3. There is concern that magnetic interference would cause problems.

[osrf-migration]

Original comment by John Hsu (Bitbucket: hsu, GitHub: hsu).

---

to add to @nkoenig, we did quite a bit of in house interference testing with Ripple sensors, and found the interference to be a non-issue.

[osrf-migration]

Original comment by Michael Wolfson (Bitbucket: nosflowm).

---

Left-handed users will be fully supported in a later release. Do you expect to be frequently changing between left- and right-handed users? If so, how cumbersome is it to move the camera and markers?

As for OptiTrack vs. Polhemus, the decision came primarily down to early feedback and concerns about magnetic interference. Doug has seen this factor in his own research efforts. We have found that the OptiTrack does indeed have some issues with occlusion, but by carefully selecting your workspace and environment, it should be capable of achieving HAPTIX goals (until the actual limbs are available to teams). I would suggest that you take some time to carefully align the camera and workspace to ensure that a nearly-complete range of motion is available. You can contact Justin or Heather for tips and tricks.

[osrf-migration]

Original comment by David Kluger (Bitbucket: DKluger).

---

We do not know how frequently we will be changing between left- and right-handed patients just yet. The camera should be easy enough to readjust. The ease of configuring Motive to accommodate for left hand, right hand, and head motion tracking won't be clear until this functionality becomes available.

[osrf-migration]

Original comment by Steve Peters (Bitbucket: Steven Peters, GitHub: scpeters).

---

• changed priority from "major" to "minor"