Boston Dynamics

CSGnet Archive CSG2018
1

[Rupert Young (2018.07.06 15.15)]

  I've long wondered how BD go about things, but have been unable

to find out as they don’t publish much on their website. So I
thought I’d look at their patents, which are quite revealing.

Here’s one,

  Looking at figure 5 it does look to me like a control of output

approach, whereby transfer functions need to be defined in order
to determine outputs to be applied, such as force. PCT doesn’t
need this (complex) transfer function stage.

Here’s a longer list of their patents

···

https://patentimages.storage.googleapis.com/ed/32/9e/993875e41a668c/US9925667.pdf
https://patents.justia.com/assignee/boston-dynamics-inc

Regards,
Rupert

2

[Bruce Nevin 2018-07-16_14:11:14 ET]

Bravo Rupert!

The ‘appendage’ is specified as having one of two states, swing or stance.

This one is especially clearly about planning movements:

https://patents.justia.com/patent/9975245

I judge from this that the environments in their videos are more constrained than they appear to be–a steady slope, for example.

···

On Mon, Jul 16, 2018 at 10:16 AM Rupert Young csgnet@lists.illinois.edu wrote:

[Rupert Young (2018.07.06 15.15)]

  I've long wondered how BD go about things, but have been unable

to find out as they don’t publish much on their website. So I
thought I’d look at their patents, which are quite revealing.

Here’s one,
https://patentimages.storage.googleapis.com/ed/32/9e/993875e41a668c/US9925667.pdf

  Looking at figure 5 it does look to me like a control of output

approach, whereby transfer functions need to be defined in order
to determine outputs to be applied, such as force. PCT doesn’t
need this (complex) transfer function stage.

  Here's a longer list of their patents

https://patents.justia.com/assignee/boston-dynamics-inc

–
Regards,
Rupert