In this paper, the authors presented the system architecture of their ground station AWE testbed, followed by some practical discussions on tuning the controller to track a figure of eight motion. The paper is well-written and provides some interesting insights for further research.

MINOR COMMENTS

1. The discussions surrounding the individual elements of each flight (lines 251-265) are a bit hard to follow. This makes it difficult to follow subsequent discussions, which are based on lines 251-265. I suggest keeping figure 5 but adding two more annotated figures for flights A and B, then expanding lines 251-265 to describe the components of flight A and flight B trajectories separately.

2. The authors mentioned the addition of a third tether line for pitch control, but there's not much (if any) discussion on the improved pitch dynamics. I assume that the window to do pitch dynamics flight tests has passed. If so, some comments on how the third line can improve flying qualities and/or direction for future studies regarding this would be appreciated.

TECHNICAL COMMENTS

Some suggested typo/grammar fixes:

1. Line 83: we refer to CL and CR as the points where

2. Line 134: please format the reference

3. Line 291: same order of magnitude as

4. Line 308: and as noted in the first paragraph

5. Line 321: double use if it/its. Unsure which one is which.

6. Line 331: Young's

Please see comments in the attached pdf file.

=============

Summary

=============

The paper presents the significant extension of an existing small-scale test bed for airborne wind energy.

The choice of wing structure is a rigid-framed delta kite, a concept which is also being pursued by the copmany Enerkite, one of the more advanced AWE companies. The kite is controlled from a ground station via three lines, allowing steering and pitching of the wing and reeling in and out. Next to the advancements with respect to hardware (adding the third line, line tensionmeters and load cells), the main contribution is the development and experimental validation of a PID-controller based on a mixed waypoint/continuous reference path. The flight experiment data are then utilized to develop models for some of the system subcomponents.

The paper is written very well and provides an extensive overview of the existing literature and state-of-the-art. Overall, it is a novel and interesting addition to the existing body of experimental AWE work and can provide a useful resource for practitioners and researchers alike.

===============

Minor comments:

===============

- Some more context could be provided to motivate the choice of control strategy. Why was this specific strategy chosen, what are its advantages and limitations, in particular in comparison with existing (published) strategies for delta kites?

- While the third kite line for pitch control (as well as the option for reeling in and out) is mentioned as a contribution of the paper, I could not find a corresponding block in the control block diagram (Fig. 4). It would therefore be useful to devote a paragraph outlining a possible extension of the controller to allow for these extra built-in features of the hardware setup.