Captain’s Log 08/02 – 08/27

August 2, 2021

The boat is finally back on track after floating around for about a week without any good wind. Now that the wind has picked up again, the boat is sailing (albeit slowly at .33 miles per hour) on an almost direct westward heading. But just like the literal calm before the storm, it looks like LoCARB may encounter its first real storm within a week.

With the current heading of the storm as shown in this animated image above, it looks like LoCARB will hit some rough conditions sometime Friday. I wanted LoCARB to experience at least one storm just to see how durable the boat really is, and also to see how the boat behaves with the winds. Will the storm drag the boat with it? Will the boat just go silent? Once the boat begins to feel the effects of the wind, Ill shorten the update interval to see how it fares. Exciting!

Also confirmed testing earlier in the week, that both motors are completely non-operational. The boat shows no directional change when the backup thruster is powered, nor is there any indication of current consumption. However, the battery voltage peaks (I saw 12.7v!) while motor is powered on. So that means there must be an additional current flow/path/drain/short somewhere in order for the voltage to bump up while solar charging. Unsure what it is, but it’s interesting…


August 4, 2021

LoCARB has lost GPS lock again after about 9-10 days (again). Power cycled, motor power off, interval update set to 8 hours.

I am waiting for the day where GPS just fails to function. Its not a huge deal though since I have a backup albeit somewhat inaccurate GPS source in the Rockblock satellite modem.

Rockblock satellite modem provided GPS location (in tracker map) 


August 7, 2021

When I launched LoCARB I had integrated a primary thruster (which I assumed had an estimated maximum life of three weeks), and a backup underwater thruster for propulsion. Both motors have since gone kaput, however…unbeknownst to me at the time, the design of LoCARB itself provided another means of propulsion.

LoCARB has been undeniably sailing for the past two weeks.

At first I believed that maybe it was just drifting with the winds, but after observing the behavior of the boat, it is most definitely steering towards its waypoints despite the southern wind and current direction.

After some simple tests (placing the waypoints closer together) I found that by setting the waypoints almost even with the boat longitudinally, the boat tends to track its heading westward in a more pronounced manner. It’s sailing!

It could be that the solar panel platform with its many honeycomb-like orifices, or the big brain box protruding above the main body provides enough wind capture to propel the boat enough for the rudder to steer the boat. It is however slow going at about .5 nautical miles per hour, you could walk faster than the boat!

At the time of this writing the boat is now 940 regular miles from launch and about 970 miles from the nearest shore of Baja California. It has traversed a total of 1231 miles of ocean (including going backwards and drifting in all the wacky directions).

On a more negative note, LoCARB’s motor pod is now at 79% humidity (launch humidity was 40%). The leak sensor hasn’t detected any large amount of water however, so there’s that. I do believe that its only a matter of time before the cells and BMS fully short circuit from moisture. How long from now that will happen? Will it last another day, week, month?

Also, the storm observed earlier this week didn’t materialize and instead, passed by harmlessly south from the boat.


August 10, 2021

3:55 pm: Reboot autopilot/Arduino, updated with new waypoint, motor off and update interval set to 8hr.


August 12, 2021

4:22 pm: LoCARB has traveled 1000 miles from launch! What a milestone! I never imagined the boat would survive this long, and performing so well despite all the malfunctions. I really consider this experiment a success, and being able to watch its journey these past 10 weeks has been a blast. It almost makes the headache in building it, worth it!

The boat is actually at 999.10 miles from launch at the 4:22 pm update.


August 16, 2021

5:13 pm: Reboot autopilot/Arduino, updated with new waypoint, motor off and update interval set to 6hr.


August 24, 2021

2:40 pm: Right on schedule, the Autopilot lost its brains and needs another reboot. Reboot autopilot/Arduino, updated with new waypoint, motor off and update interval set to 8hr.


August 25, 2021

5:50 pm: A little bit of a scare yesterday and today…LoCARB did NOT respond to the set of commands I sent it yesterday. It continued to report back every 8 hours (which meant it did not reboot since after a reboot the update interval reverts back to 30 minutes), and even did an update 3 hours (at 9:35am) after the 6:39am update (3 hours is not a selectable interval). This means that the Arduino probably has/had a glitch in keeping time after running uninterrupted over a week. As to why it does not process incoming satellite messages though, I am at a loss. The Arduino should process the incoming message immediately after a successful outgoing send no matter what timing is involved.

I’ve seen this time glitch once before and thought it was strange, but I never tried to send it a command prior to the time glitch. Because Arduino millis() does rollover and reset to 0 once it reaches a maximum value, I thought maybe it could be that, but the rollover takes about 49 days with no reboot to happen. I’m not sure what the issue is but, as long as it corrects itself afterwards its not a big deal. I just have to expect the behavior after a week of being continuously powered on.

Either way, at 5:35pm, the Arduino power cycled both itself and the autopilot (which is good because the autopilot was having trouble with rudder control and voltage monitoring), and will update again in 30 minutes (after the initial boot update). For troubleshooting purposes I’ve sent a command to update the interval to 1 hour and turned off the motors. If this takes, ill switch to 8 hour update intervals.

LoCARB is about to hit some high winds come tomorrow. Looking forward to see how the boat handles it.

6:16 pm: LoCARB is operating as it should. On a very mind blowing find…after checking the 30 minute update, I realized that the primary RC motor fully submerged in salt water STILL WORKS (after a reboot the motor defaults to ON and throttle 2)! The boat reports back the correct target RPM reading. This means that cheap RC motor is functional after 12 weeks fully submerged in salt water with about 6 weeks running and 6 weeks turned off. I would assume that after being stopped for 6 weeks, the motor would seize from corrosion, but that’s not the case.

For the most part however, LoCARB electronics have been very reliable and working together quite nicely. I am very pleased considering I’ve never built anything this sophisticated before 🙂

7:16 pm: Update Interval set to 8 hours. I did not update the boat with waypoint 11 yet.


August 27, 2021

Thursday was probably the most eventful day LoCARB has had in a while. Not only did a moderate weather event pass directly overhead with wind speeds fast enough to propel the boat so it could steer northwesterly, and strong enough to trigger a motion event, but the boat also finally registered a reading on the leak sensor (from the water sloshing all over from the turbulence).

Here you can see the storm passing overhead with winds 23kts and higher


You can see the leak level in battery pod shows a 6. This is probably around 1″ of water around the sensor (from pre-launch testing). There is a 3d printed barrier which was used to keep the battery from sliding back and forth on each end of the battery pack which may be keeping water from contacting the leak sensor on a regular basis (there are openings halfway up the barrier). I believe the weather conditions were so rough that the boat actually capsized (or almost did) and water from the rear of the pod finally went up and over the barrier to fill the front end of the compartment (where the leak sensor lives).

In the next update however, there was no reading from the sensor which means water was not in current contact with the sensor (could the update have caught the boat in just the right tilt position where most water was accumulated against the leak sensor?). Either way, both a reading of the leak sensor and pod humidity sensor confirm liquid water within the submerged pod enclosure. I also realized that this is the first time every sensor integrated on the boat has been used to get an idea of the condition the boat is in.

When the boat hits high winds, the force is actually enough for the boat to steer itself. In the image above you can definitely see the boat trying to head north west to its waypoint. The last marker shows the position when the winds have dropped back to around 14-17kts.