I started by creating a GIF of the video reference that I found (front reference and side reference) and then I made frames from the GIF using the GIF Scrubber app for Chrome. This created 70 frames from my 6.4 second reference. I took every 8th pose and posed my model in Maya the same way. I forgot the do the math and finished on frame 70 in Maya but of course I should have check the length of my reference first and realised that I should have had around 154 frames. Once I had the poses on the correct frames I went back and started to adjust the timing between each pose.
This is my animation before I added the box to the action:
Jill Box Lift (Without Box) Side View:
Jill Box Lift (Without Box) Front View:
This is where I tried including the box with the action. Of course, the arm spacing to begin with wasn’t correct to hold the volume of the box the whole way throughout, so I had to go back and tweak the translate/rotation values for the shoulders, elbows and arms. This was quite difficult as the hands would often move away from the box between the poses so I ended up setting quite a few keyframes. There is still a lot of slipping and sliding that I need to fix and I think that I could also work on the weight a lot more. For example, there could be more anticipation/gathering of energy before she lifts the box above head height.
Jill Box Lift Front View:
Jill Box Lift Side View: