Hi all,
We’re a hang gliding + paragliding club and we operate an electric towing winch on a 1200 m grass airstrip. The winch has two drums and is remote-controlled, so no one is standing at the winch during operation.
After two tows, we need to retrieve the tow ropes using a car. To reduce downtime between tows, we’d like to build an autonomous rover that can:
-
Drive from the winch/start area to the far end of the strip (take-off area)
-
Pick up the tow ropes there (mechanical gripping solution will be developed later)
-
Return and drop off the ropes at the winch/start area again
I’ve attached a simple sketch of the layout (winch at one end, rover route along the strip).
Environment
-
Grass field / airstrip, ~1200 m long
-
Mostly flat, but can be uneven / bumpy
-
Outdoor conditions: wind, rain/humidity, temperature changes
Performance requirements
-
Cruise speed to avoid delaying operations: ~15–20 km/h
-
Accuracy while driving along the strip: ~1 m is OK
-
High accuracy only near the winch / rope pickup/drop-off point: ~2–5 cm desired
- The rover can slow down significantly for the final approach / docking maneuver
Current hardware concept
We’re considering using a large 4WD RC car platform (possibly with extra battery capacity), e.g. Kraken KV5TTE Ultra 2 Electric (KT7707) or similar.
For the autopilot/navigation/telemetry stack we’re currently looking at:
-
The Cube Orange+ autopilot set (ArduPilot Rover)
-
Holybro PM07 V2 power module (14S capable)
-
ArduSimple long-range RTK base + rover GNSS kit (cm-level RTK, incl. radio link)
-
RFD 868x telemetry radio pair (EU 868 MHz)
What we’re asking for feedback on
We’d really appreciate a candid assessment of:
-
Overall complexity & risk
- How hard is this to get reliable in real life (not just “it moves once”)?
-
Navigation / accuracy strategy
-
Is RTK alone typically sufficient to achieve 2–5 cm repeatable docking accuracy in the last meters?
-
Would you recommend adding something for the final approach (e.g., fiducial marker + camera, lidar/rangefinder, wheel odometry, etc.)?
-
-
Component choices
-
Are the Cube Orange+, PM07, RTK kit, and 868 MHz telemetry choices reasonable for this use case?
Any alternatives you’d strongly recommend (especially for GNSS/RTK robustness on a ground vehicle and for high-speed driving on grass)?
-
Notes / constraints
-
Rope gripping mechanism is not the focus right now; we mainly want to validate the feasibility and the autopilot/navigation setup.
-
The rover is intended to run along a defined corridor (the airstrip), not complex off-road navigation.
Thanks in advance for any guidance!

