Single-Message StationKeeper Incidents, Full Season

Why this report focuses on single-message encounters. Across the 2025-26 season, StationKeeper sent 11,370 speed advisories to 1,515 distinct vessels (program totals, all stations). Most vessels were messaged repeatedly as they transited an active SMA. Those repeat-message cases are hard to read cleanly: once a vessel has received several advisories, any later speed change can't be attributed to a single event. This report isolates the 700 encounters (500 distinct vessels) where a vessel received exactly one advisory and remained trackable for 20+ minutes afterward. An encounter is one message to one vessel on one transit; the same vessel can appear more than once. The cohort excludes Tybee Island area transits and pilot vessels. In that subset, 555 of 700 encounters (79.3%) were later observed at 10 kt or less (78.4% of 653 when government vessels are also excluded). Vessels with similar speeds and routes that met the alert criteria but received no message slowed at about the same rate, so this figure should not be read as an effect of the message.

Comparison with unmessaged vessels. Against matched vessels that met the alert criteria but were not messaged, the difference in reaching 10 kt or less was about 0 (plus or minus 8 points, 95% interval). The results are the same when government vessels are excluded. We found no measurable added slowing from the message.

How to read this number, and one important limitation. The 79.3% describes what happened within this single-message subset; it is not the share of all messaged vessels that slow, and it is not a measured effect of the messaging. Two caveats travel with it:

1. Selection on response. Isolating single-message encounters is not a neutral filter. As our season-baseline analysis states in its own words, “restricting to single-message transits selects on response, non-responders draw a second message and leave the population.” A vessel that keeps overspeeding tends to trigger a follow-up advisory and is thereby moved out of this single-message group, so the subset is weighted toward vessels that had already begun to respond. The true share of vessels that slow after a first message is lower than 79.3%.

2. Descriptive, not causal. This is a record of what vessels did after a message, not proof the message caused it, many vessels are slowing on pilot advice, at zone entry, or for traffic regardless of our alert. Isolating a messaging effect requires a control group; the comparison above summarizes that analysis, and the season-baseline analysis shows the ambient corridor speeds these encounters should be read against.

Resolved vs. Gap. “Resolved” marks an encounter where the vessel was observed at a compliant speed (≤10 kt, per 50 CFR 224.105) in its post-message track; “Gap” marks one where that compliant observation is not present in the trackable window. Exact classification thresholds are defined in the investigation report.