Pi YAPP Anchor Alarm

[gregcope]

Hi All

What is a reasonable Anchor alarm radius?

For example I have a GPS, that appears to have a GPS EP +/- of anything from 2M to 18M or more.

What would a good / reasonable anchor alarm radius be (I know people will want 0... )?

Beyond what range it is not useful? 40M, 80m, 400M? I know it depends, but thinking of average use cases. Like anchoring at East head for the night.

My Raspberry PI boatmon solution YAPP is coming on. Code complete. Mostly testing now with tweaks/fixes, whilst waiting for parts to arrive;

Here is an iPhone screen short of setting an anchor alarm via SMS and the answer back;
View attachment 47492

Code / some basic doco here; https://github.com/gregcope/rpi

I am waiting for the gromets from China, and then I plan to raid the tupperware draw and do mobile testing. Also need to test the Bilge Switch logic/switch.

 

[Joker]

There's also the issue of the boat swinging round the cable. Mine sheers from side to side like nobody's business.

 

[laika]

This was actually a great question with relevance for anyone *using* a bought anchor alarm app as well as anyone writing it. As I don't claim to have the answer...**bump**

I think you'll need any app to be able to put in a distance and bearing to the centre of your alarm circle: the point you feel comfortable swinging around, possibly the estimated anchor position. I think some apps have a button to press when the owner thinks the anchor has hit the floor but generally I'm too busy anchoring to bother with that.

You need to account for distance of the GPS sensor to the bow, bow to sea, amount of chain out and tidal height. Extra credit for having variable alarm circle according to state of tide but having done some finger in the air calculations for east head, given the GPS variation that might be more hassle than it's worth.

I'd argue that this is something you might like to start conservative with, accept false positives, and expand the circle as experience dictates. Then post back here obviously . Does GPS error tend to stay pretty constant in one location over the course of an evening or does it vary with each reading? If the latter, a smoothing algorithm presumably is needed...

 

[OceanSprint]

I know nothing about it, but could the app average over time, then alert if position went outside a certain numberof standard deviations of the average? But with maximum distances set as well. But would need quite a bit of experimentation at anchor under different circumstances.

 

[duncan99210]

Anchor alarm requirements:

1. Ability to set where the anchor is by reference to a bearing and distance from the alarms current location on the boat.

2. A swinging radius to accommodate wind and tide changes.

3. An ability to define an arc within the swinging circle reflecting the need to avoid shallow bits.

4. GPS averaging over say ten readings before alarm sounds. This will help prevent false alarms due to minor wobbles in the GPS signal.

Most GPS these days will give a consistent position +/- 5 or so, so that would be a starting point for the alarm to work to.

 

[gregcope]

Anchor alarm requirements:

1. Ability to set where the anchor is by reference to a bearing and distance from the alarms current location on the boat.

Tricky - will have to look up the maths for that.

2. A swinging radius to accommodate wind and tide changes.

Done

3. An ability to define an arc within the swinging circle reflecting the need to avoid shallow bits.

Ah. Might be tricky.

4. GPS averaging over say ten readings before alarm sounds. This will help prevent false alarms due to minor wobbles in the GPS signal.

Done.

Most GPS these days will give a consistent position +/- 5 or so, so that would be a starting point for the alarm to work to.