|
|
08-10-2017, 08:20
|
#61
|
Registered User
Join Date: Mar 2012
Location: Boston, MA
Boat: 1981 Bristol 32 Sloop
Posts: 17,994
|
Re: Autopilot - Open Source
champ
THIS IS EXACTLY MY POINT. (You need to read the thread.)
WE DON'T WANT OR NEED Bathymetric!!
With Bathymetric the topgraphic "BOTTOM" is what they want so they have to have higher power LIDAR (green) and run the planes lower with a tighter pattern because they are trying to get a "return" from the GROUND which is many meters below the WATER SURFACE.
We are interested in the WATER SURFACE, an entirely different problem. Perhaps you can advise what you know about that? We have a specific application and a small Radar may be better/more cost effective.
Quote:
Originally Posted by champ8242
Bathymetric liDAR systems are far more complex than topographic. Green instead of red laser, they scan slow to get the power they need and the processing is much more complex. This adds up to systems that cost a lot to run. On top of that, water clarity is a huge issue. If your signal is not strong enough you can’t see it. You can use a lot of time and money sitting on the ground waiting for clearer water. This makes bathymetric LiDAR expensive relative to topographic LiDAR.
|
|
|
|
08-10-2017, 08:54
|
#62
|
Registered User
Join Date: Oct 2009
Location: Lelystad, The Netherlands
Boat: vd Stadt 44, Ben the Boat
Posts: 238
|
Re: Autopilot - Open Source
Quote:
Originally Posted by rgleason
champ
THIS IS EXACTLY MY POINT. (You need to read the thread.)
WE DON'T WANT OR NEED Bathymetric!!
With Bathymetric the topgraphic "BOTTOM" is what they want so they have to have higher power LIDAR (green) and run the planes lower with a tighter pattern because they are trying to get a "return" from the GROUND which is many meters below the WATER SURFACE.
We are interested in the WATER SURFACE, an entirely different problem. Perhaps you can advise what you know about that? We have a specific application and a small Radar may be better/more cost effective.
|
THAT WAS EXACTLY MY POINT. THERE NO USE OF LiDAR ON WATER
You want to look horizontally, with bathymetric you would look vertically!!
|
|
|
08-10-2017, 09:49
|
#63
|
Registered User
Join Date: Mar 2012
Location: Boston, MA
Boat: 1981 Bristol 32 Sloop
Posts: 17,994
|
Re: Autopilot - Open Source
Champ
I am trying to understand this better. Your post Pier of Scheveningen.
3dgisviewer.eu (nice work) shows a definite "Water Surface" returns down the swath of LiDAR along the pier. Additionally when I rotate the 3D to look in section the returns do not show lower than the surface (which is perfect for our use for determining the Water Surface). Was the lidar data set extended further than what is shown? If so, was it just not used for the graphic or were there absolutely no returns from the Water Surface areas further from the pier?
Quote:
Originally Posted by champ8242
You can forget LiDAR. It wil not penetrate water and you wont get returns.
I can show you a project I few with LiDAR over the pier of Scheveningen.
3dgisviewer.eu, first project.
You can see water on the sides but mostly some water patches on the pier that doesn't give returns of the surface. btw, the point cloud is clororized. This pretty picture you wont get from just LiDAR
|
|
|
|
08-10-2017, 10:52
|
#64
|
Registered User
Join Date: Oct 2009
Location: Lelystad, The Netherlands
Boat: vd Stadt 44, Ben the Boat
Posts: 238
|
Re: Autopilot - Open Source
LiDAR (active sensor) creates a point cloud, meaning a cloud of points. You get a point by sending a laser light and wait for a return. No return, no point. So far simple! So what is a point, simple as well! Lat, long and Alt. This is very simplified, there is more, but in essence, where does the point (the return) lives in space.
Once we get a return, speed of light, time divided by 2 we will know a distance. Now with distance we know one of the parameters, but we need to know where the laser light was pointing to. With LiDAR we use an IMU for this. With the IMU’s known direction of the beam and distance and “our” known point, we can calculate the coordinates.
BTW, you can get laser devices from a DIY market. Search the web for DIY laser measuring.
A LiDAR sensor might consists of 16 or 32 or 64 laser heads generating an N-number of laser beams / X00.000 measurements a second. Mentioned in this thread, a Velodyne VLP16, creates 300.000 beams a second (360 degrees). This means could do 300.000 measurements a second. Only and only if it gets a return.
So, if you make the right calculations, per return you know where the laser was pointing at, you can get the coordinates of the return. Now if you make the point visible, give it an RGB color, you can display it.
I made a few weeks ago a youtube video, Here you see the laser points displayed, real time, giving a color depending on the height of the object.
Now, a) no return, no point! Water doesn’t return the light so no points. B) @ 300.000 beams a second being points, with a minimum of lat, long and alt values, imagine the size of dataset you get when you have returns.
So the water you see in the point cloud are none existing points, no color
|
|
|
08-10-2017, 11:06
|
#65
|
Registered User
Join Date: Oct 2009
Location: Lelystad, The Netherlands
Boat: vd Stadt 44, Ben the Boat
Posts: 238
|
Re: Autopilot - Open Source
PS,
Attached, a snapshot of the point cloud, to see the returns
|
|
|
08-10-2017, 13:25
|
#66
|
Registered User
Join Date: Mar 2017
Posts: 804
|
Re: Autopilot - Open Source
Quote:
Originally Posted by champ8242
You want to by the sensor? € 89,000.00 but has less than 2cm accuracy!
|
LOL We all agree LIDAR is not suitable (not for the same reasons perhaps ), but who gives a monkey's whatsit about 2cm accuracy... 2 foot accuracy on the position of a wave would be plenty of accuracy for this purpose.
|
|
|
08-10-2017, 15:19
|
#67
|
Registered User
Join Date: Aug 2009
Location: oriental
Boat: crowther trimaran 33
Posts: 4,449
|
Re: Autopilot - Open Source
I think the starting point is using a regular camera in the daytime...
|
|
|
08-10-2017, 20:32
|
#68
|
Registered User
Join Date: Mar 2012
Location: Boston, MA
Boat: 1981 Bristol 32 Sloop
Posts: 17,994
|
Got it. LiDAR is not the answer.
Champ8242
Quote:
So the water you see in the point cloud are none existing points, no color
|
So the laser pulses do not bounce off the water at all and the blue areas that I saw along the pier are actually "GROUND"!! Sorry about that, but these things do happen.
Thank you for taking the time to explain this to us. I had hoped that lidar would be right this application, I guess its camera and then maybe mini radars!
|
|
|
09-10-2017, 00:23
|
#69
|
Registered User
Join Date: Mar 2017
Posts: 804
|
Re: Autopilot - Open Source
Quote:
Originally Posted by boat_alexandra
I think the starting point is using a regular camera in the daytime...
|
Agreed. Once the system can steer around waves detected by any source is shouldn't be much of a change to add in another source.
In the meantime I'm going to mess around with some detection ideas. It may take some time!
|
|
|
09-10-2017, 04:03
|
#70
|
Registered User
Join Date: Oct 2009
Location: Lelystad, The Netherlands
Boat: vd Stadt 44, Ben the Boat
Posts: 238
|
Re: Got it. LiDAR is not the answer.
Quote:
Originally Posted by rgleason
Champ8242
So the laser pulses do not bounce off the water at all and the blue areas that I saw along the pier are actually "GROUND"!!
|
The acquisition was done at low tide, it's the sand in salt water.
|
|
|
09-10-2017, 09:19
|
#71
|
Registered User
Join Date: Jan 2010
Location: Harlingen, NL
Boat: KMY Stadtship 56
Posts: 528
|
Re: Autopilot - Open Source
A Dutch engineer is going to take his predicting waves with radar research and taking it commercial.
Next Ocean | Less Risk, More Uptime
They use the raw data stream from a normal radar, probably with sea state tuned to a different level than normal.
|
|
|
09-10-2017, 13:50
|
#72
|
Registered User
Join Date: Mar 2012
Location: Boston, MA
Boat: 1981 Bristol 32 Sloop
Posts: 17,994
|
Re: Autopilot - Open Source
Thanks Merrimac. It looks like Radar is the choice.
Next Ocean | Less Risk, More Uptime The click on Technology.
Quote:
The main input for Next Ocean’s wave prediction system comes from the ship’s navigation radar. An additional dedicated radar antenna is therefore not required.
Using established technology the directional wave spectrum and surface current are derived from the raw radar data. In addition to this technology, Next Ocean has developed new algorithms that provide valuable extra information: real-time prediction of approaching waves and the resulting ship motions minutes in advance.
The system is currently undergoing operational tests in the North Sea. Test results are expected by the end of 2017. The Next Ocean wave prediction system is targeted for market entry in 2018.
|
Wave Predictor
I wonder if this couldn't be "Wave Avoider", since in our case the vessel is moving...
There are lots of ways to avoid a particular spot at a particular time if you can predict when and where the wave will be.
|
|
|
09-10-2017, 18:27
|
#73
|
Registered User
Join Date: Aug 2009
Location: between the devil and the deep blue sea
Boat: a sailing boat
Posts: 20,945
|
Re: Autopilot - Open Source
Given limited means, are you not better off measuring what the boat is doing and then having the AP act accordingly than looking out for an incoming wave?
How do you know exactly what the boat 'will do with the wave'.
At times a boat will spoon and veer at other times she will come over and back. This at least true for fast running or very deep broad reaching. Otherwise windward capsizes would never happen to dinghy racers.
So you may be able to spot the wave but not be able to guess in advance what the boat will do when the wave hits.
And at times the wave is there but it does not hit. If you have sailed enough, you must have seen phantom waves that are there when you look and you all duck and brace and then ... nothing happens. If a human gets fooled like that, also a chipset can be fooled.
Always mind helms were tied to a pole to avoid just that - looking back at the waves.
So my point I think is we may be overestimating the importance and usability of early information on approaching waves.
barnakiel
|
|
|
21-10-2017, 19:52
|
#74
|
Registered User
Join Date: Mar 2012
Location: Boston, MA
Boat: 1981 Bristol 32 Sloop
Posts: 17,994
|
Re: Autopilot - Open Source
It is possible there is no advantage to knowing what will arrive, but there is know way of knowing this without trying to decipher and use the information.
At this point I think Sean is thinking it would be more useful to know when there is a boat on the horizon.
|
|
|
22-10-2018, 05:45
|
#75
|
Registered User
Join Date: Apr 2017
Location: IJmuiden, NL
Boat: Beneteau Oceanis 500 Clipper
Posts: 59
|
Waves and AP
You all seem to be looking above water. Is it not simpler to look under water? After all, waves have - I gather - a profound effect on speed and direction of the under water flow.
Could not that be measured? Maybe something like 'Faster flow means bigger wave' os something like that?
No clue if there is any affordable means of measuring that ... maybe we get back to sound albeit under water?
|
|
|
|
|
Thread Tools |
Search this Thread |
|
|
Display Modes |
Rate This Thread |
Linear Mode
|
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
|
Advertise Here
Recent Discussions |
|
|
|
|
|
|
|
|
|
|
|
|
Vendor Spotlight |
|
|
|
|
|