When Your Satellite Window Is 4 Minutes
The Navy's unmanned vessel program had a problem: their ML models needed updates, but the vessels were underwater. Satellite links cost $14/minute and dropped constantly. Cloud-based ML wasn't going to work.
Evidence boundaryHistorical case-specific report. Figures describe the deployment documented at the time and are not current universal product guarantees.
Client
U.S. Navy
Industry
Products
Expanso, Mycelial Kafka Connector
Timeline
Initial vessel in 11 weeks, fleet rollout ongoing
ROI
Eliminated dependency on real-time connectivity
/ Challenge
The Challenge
The unmanned vessel program was stuck. ML models trained on shore worked great - until you tried to update them on vessels operating in the Western Pacific. The existing approach required vessels to surface, establish satellite link, and maintain connection for the full update. Success rate was around 23%.
Satellite bandwidth runs $14/minute and connections drop mid-transfer
Vessels surface for 4-6 minutes on average before diving again
Failed updates meant vessels ran stale models for weeks
DoD security review for any new software takes 9 months minimum
Existing ML platform required constant cloud connectivity
Program was 14 months behind schedule on autonomy milestones
/ Solution
The Solution
We built a system that assumes the connection will fail. Model updates break into small chunks. Vessels grab what they can in each window. The orchestrator tracks what each vessel has and what it still needs. A full model update completes across 3-4 surface windows instead of requiring one long session.
Chunked Model Delivery
Model updates split into 200KB segments. Each chunk verifies independently. Vessels resume from last successful chunk - no wasted bandwidth on retransmission.
Fleet-Wide State Tracking
Shore command sees exactly which models each vessel has, when they last connected, and what updates are queued. Priority vessels get updated first.
Opportunistic Data Return
Vessels collect sensor data continuously. When they surface, high-priority data uploads first. Raw feeds compress and transfer during longer windows. Nothing gets lost.
/ Results
The Results
The program caught up on its autonomy milestones. The historical report records model update success increasing from 23% to 97% under intermittent connectivity.
97%
Update Success
72 hrs
Fleet Propagation
11 weeks
First Vessel
0
Failed Transfers
Model update success rate jumped from 23% to 97%
Full fleet receives updates within 72 hours of release
First operational vessel deployed in 11 weeks
Satellite costs dropped 34% - less retransmission, smaller payloads
Vessels now run 3 concurrent ML models instead of 1
Program recovered 14-month schedule slip in 6 months
/ Explore Related Solutions
/ What's next
Deploying to disconnected environments?
If your edge devices cannot maintain constant connectivity, we can evaluate chunking, resumability, and prioritized transfer against your operating constraints.
/ Typical engagement
30 days
Free assessment
90 days
Optional pilot
Measured
Baseline vs. after
