To make this website work, we log user data. By using Shephard's online services, you agree to our Privacy Policy, including cookie policy.

Open menu Search

How adverse weather conditions affect autonomous USV navigation

14th June 2024 - 08:50 GMT | by Alix Valenti


USVs Seahawk and Ranger transit the smooth waters of Sydney Harbour during last year’s Integrated Battle Problem 23.2. Higher sea states have been causing problems for the systems’ autonomy however. (Photo: US Navy)

Significant advancements and ongoing challenges in the development of autonomous navigation for uncrewed systems have highlighted the complex interplay between human oversight and technological capability.

US Naval Sea Systems Command (NAVSEA) recently provided an update on its uncrewed maritime systems. Along with detailing the progress of each individual programme, NAVSEA shared the latest results of its autonomous navigation testing on USV prototypes.

This aspect was particularly intriguing as it has prompted a discussion on the definition of autonomous navigation.

“One of the goals of prototyping is putting the vessel in as many diverse situations as possible to explore where the boundaries are and in doing so find the risks associated with unmanned operations,” Capt Scot Searles, programme manager – unmanned maritime systems for NAVSEA.

Are ‘kamikaze’ USVs a game changer in naval warfare?

Turkish MARLIN USV operates alongside manned units for the first time

Exail completes trial of new large long-endurance USV

The US Navy (USN) has been testing its large USV prototypes for the past two years, accumulating nearly 10,000 hours of autonomy data across more than 100,000nmi of travel. During those tests, a crew was posted on board the USVs to monitor the craft and, if necessary, intervene.

Searles prefaced his comments by stating: “Human captains can be over-cautious; during Integrated Battle Problem [IBP] 23.2, the prototype vessels were disabled 13 times by the vessel master, that is once every eight days, but less than half were described as actual concerns.”

What “actual concern” really means, probably, is in the eye of the beholder.

IBP is a multi-domain uncrewed capabilities exercise that falls under the US Pacific Fleet’s Experimentation Plan. The 23.2 iteration concluded in January 2024 and involved four USVs (Mariner, Ranger, Seahawk and Sea Hunter) which travelled a combined 46,651nmi to make port visits to Hawaii’s Pearl Harbor, Papua New Guinea, Yokosuka (Japan) and Sydney (Australia).

Searles proceeded to explain that during IBP 23.2, the overly cautious crew intervened 157 distinct times – that is, once every 28 hours. According to the programme manager, more than two-thirds of these interventions were unrelated to autonomy issues. Rather, they traced back to problems with Government Furnished Equipment C4I systems (reloading crypto or losing crypto synchronisation) or were done for operator convenience.

This leaves 48 human interventions, including 11 related to hull, mechanical and electrical (HM&E) issues, three due to weather events, three that could not be categorised, and 17 for sensing and perception, which were, according to the captain, highly dependent upon weather.

Only nine of the 17 interventions of concern were related to the manoeuvring decisions of the autonomy itself, Searles proudly concluded.

Now, while there is no doubt that the USN is making significant progress with its USVs, it might be a little concerning that it does not count issues with sensing and perception in bad weather (i.e., chasing false contacts) as causes for concern over its systems’ autonomy.

Where will those systems operate? In a calm bay somewhere where the sun always shines? Don’t two rather rough oceans flank the American continent?

Autonomous navigation is not only defined as a system’s ability to avoid collisions – something the USN’s USVs are apparently now handling well, according to Searles – but also by its ability to navigate in different sea states.

Admittedly, most industry experts say these systems will probably not be used beyond Sea State 3 (large wavelets, crests beginning to break, scattered whitecaps). Many of those same experts, however, affirm that current autonomy algorithms still struggle to distinguish between whitecaps and actual obstacles.

So, it seems that the human intervention level on USN USVs, currently, is once every four days – and not 12 – because bad weather counts.

In the end. USN crews are not being overly cautious; they are doing their job. And that is fine. There are still many nautical miles to cover before achieving complete autonomy (and crew trust). There is no shame in admitting that.

This analysis article originally appeared in April's Decisive Edge Naval Newsletter. To receive regular updates from Alix Valenti and our team of defence experts visit our Decisive Edge sign up page.

Shephard's Eurosatory 2024 coverage is sponsored by:

BAE Systems
Alix Valenti


Alix Valenti

Dr Alix Valenti is an international freelance defence journalist. Her main focus is on naval …

Read full bio

Share to