Trial executed by the Royal Dutch Coast Guard (KNRM)

Date:​ Jan 21 2017.
Location:​ Scheveningen, Netherlands.
Water temp:​ 7°C.
Time to hypothermia:​ 6 minutes.
Weather conditions: ​1c Clear, 3 BFT, 0.3m wave height.
Sea current:​ North, approx. 1 knot.
Rescue boat:​ 65 foot aluminium, Habbeke BN133 “Valentijn”.
Test observer:​ Henk Kok.

The PanPan team arrived at 9:00 AM at the KNRM station in Scheveningen.

The team was introduced to the KNRM volunteers and quickly set up several smart devices provided by the KRNM with the CrewWatcher app. Namely a: iPhone 6, Nexus 6 and an iPad Air 2. Ultimately the Nexus and iPad would be used alongside PanPan’s iPhone 7.
There were six beacons available for the test so two beacons were connected to each individual device.

We set out at around 10:00 AM, leaving the harbor of Scheveningen with a crew of approx. ten including the PanPan team.
Five beacons got divided over the crew while one beacon was attached to the designated MOB; a man sized and weighted simulation doll, wearing an immersion suit and swimming - life vest. In total the simulation doll went overboard six times of which four were logged.

After getting some distance from the shore testing began. The simulation doll would be thrown overboard at the stern at an unknown time to the cabin crew. The first test was conducted at a speed of 6 knots with the test observer at the stern deck. Once the MOB occurred the CrewWatcher app alerted the cabin crew most notably not by the audio alarm (which was drowned out by the combined 2400BHP engines) but by the visual change in color from green to orange. The coxswain then steered for the rescue course provided by the app. The course proved to be accurate while the MOB remained in sight at all times. The simulation doll was recovered within 3 minutes on the first test.

Between tests the crew was asked to walk around the boat from stern to bow and back to test the signal reach of the beacons. Bluetooth typically has a very hard time with metals and (such as the aluminum cabin) and large distances. The CrewWatcher beacons however held their own and have not reported a single false alarm on either of the three testing devices over the course of the two hour testing period.

Several other MOB simulations were conducted at various speeds from 2 to 30 knots.
In several tests the alarm was triggered prematurely by water as the MOB was placed on the open stern gate. As the boat picked up speed “white wash” would flood the simulation doll and the alarm would be triggered. While not ideal in this testing scenario the product did work as intended. The feature was designed to alarm in case a sailor is to be dragged on by his lifeline. (also see:

To remedy this situation a new beacon, placed in a waterproof pouch was used for further testing. For actual use it’s recommended to wear the beacon inside your lifejacket.

A deviation was found between the MOB rescue location and the point of loss indicated by the CrewWatcher application. This deviation was likely caused by two things; MOB drift due to ocean current and the coxswain rightfully ignoring the app instructions to go back to the point of loss. (due to earlier visual contact with the MOB)

The calm and clear conditions of the day made it possible to establish early visual contact.
A MOB is more likely to occur in harsh conditions with poor visibility.

In such scenario the coxswain would likely return to the point of loss and follow the CrewWatcher application’s instructions to follow the sea current from there.
All the while being aided by the applications homing feature. The homing feature will show to boats location within the search perimeter and scans for the beacon’s signal to indicate close proximity to the MOB.

Over the course of the two hour trial CrewWatcher successfully alerted the crew of all MOB simulations. Average alarm time to alarm was six seconds and rescue time was always under four minutes.