/.R. Marx et al. 
1.5| 
! 
= 05 
fi) 
2 
£ 
A 
0.5 
— 
7a 
pl 
fr 
Fa 
AV DENEB 
Ocean Engineering 343 (2026) 123388 
-—SoG_ 
ref 
—5S0G 
Taf 
— 
—5S06G | 
ref‘ 
— S0G 
—r 
ref 
3) 
I 
= 
3eL4 
N 
4} 
% 
f 
Zu 
| 
Y 
—_ 
U 
‚h.. 
—=S0G 
—50G 
Trof 
n I 
_ 2] 
5 N 
3 
> 
& 
AVESSIP 
100 200 300 400 SC) 
Time [s] 
600 700 800 900 1000 
Fig. 13. Measurement data of the speed over ground and yaw rate for three vehicles DENEB, BELA and MESSIN. 
vehicle; the corresponding switch lights up blue. When permission is 
zgranted by the respective skipper or the chief mate on board DENEB, 
‘he light changes to green. The calculated evasion trajectories are send 
:O the visualization and control systems. These trajectories are provided 
with a delay to the start time, in which the ships travel directly to the 
starting point in DP mode when the MPC is active. Next, the respective 
VIPC is enabled at the touch of the button. The switch lights up red if 
chere is no valid trajectory, blue if there is a trajectory but it is too far 
away from the current pose and green if the MPC is active for the eva- 
sion trajectory. In the event of a malfunction, excessive deviations from 
the target trajectory or when the end points are reached, the captain 
declares the maneuver to be complete, external control is returned and 
the vehicles are steered manually again. 
5.2.2. Results of real-world tests 
All tests are based on the defined initial encounter situation that 
would lead to a collision. Generally before a maneuver is started, it must 
Je ensured that all system components on each individual vehicle are 
working properly. In the case of tests in the real environment, the traffic 
situation and weather conditions must be checked to see whether they 
are suitable for the defined safety aspects. On the actual readiness level, 
che fall back solution is always the manual control, ifone external option