Autopilot - Open Source

[rgleason]

To see the Lidar Data format.
Top 6 Free LiDAR Data Sources - GIS Geography

Luminar Website
https://www.luminartech.com/technology/index.html

[seandepagnier]

So try the other method. Listen to the waves, build a database and check this periodically for data drift. Then (try to) tie boat g vectors with wave train data that preceded the vectors. Now you are talking.

This sounds great but is rather vauge on how to actually implement it. I can compute fourier transform of the various inertial measurements, and this does indeed show peaks at various frequencies. These frequencies vary greatly depending on the sea-state, the boat, and the angle it is sailing.

Initially these values could be used to quickly retune the autopilot based on past experience, maybe.

Mind consequences of the mistakes such a "smart" AP will make now and then. Are they worth the potential gains.

Otherwise NKE would have been there long ago.

What are we at then?

b.

This type of logic is thrown at me constantly. It is not valid at all. The real reason NKE or whomever didn't do this, is they don't know how or have not put enough effort to succeed. It is not that it can't be done.

Consider all the features in opencpn that don't exist in any other charting program. By contrast, there are not many features missing from opencpn.

Features like climatology and celestial navigation are absent from other software. I think I can have a video camera integrated into my autopilot within 1 year in a way that improves it.

[Moo]

Originally Posted by rgleason

What about state of the sea surface?
Topobathy LiDAR – when is a good time to conduct a survey?
Large ocean waves are great for surfing, not for LiDAR!"

Also Eye Safety!!
. And I appreciate the eye-safety problem as a new laser safety officer would not let my beam-spread eye-safe laser be used because the laser INSIDE the lidar was not eye safe

Eye safe lasers are available and for the resolution required LEDs are probably adequate.

LIDAR can be used to penetrate shallow water and people are using it for inshore bathymetry already. By messing around with wavelengths it may be possible to make it see the sea surface but for the resolution and range required laser is probably overkill. Visual, acoustic ranging, or more traditional RF ranging are probably easier to implement cheaply.

Large ocean swells >10m can easily be picked up on traditional radar out to many miles but here we are looking at ranges shorter than the minimum range of traditional radar and even shorter than the new 4G radars I would think as we need to be able to watch the wave all the way up to impact with the boat.

Self drive car technology may help. I'm going to see if I can devise something with a parking radar and do some real world tests to see what I can pick up on my next voyage.

Advice and suggestions welcome

[rgleason]

Tesla Advanced Sensor Coverage

Eight surround cameras provide 360 degrees of visibility around the car at up to 250 meters of range. Twelve updated ultrasonic sensors complement this vision, allowing for detection of both hard and soft objects at nearly twice the distance of the prior system. A forward-facing radar with enhanced processing provides additional data about the world on a redundant wavelength that is able to see through heavy rain, fog, dust and even the car ahead.

[rgleason]

The Google fleet uses LIDAR (light-detection-and-ranging) technology, such as that in a system available from Velodyne’s HDL (high-definition LIDAR)-64D laser-sensor system, which uses 64 spinning lasers and then gathers 1.3 million points/sec to create a virtual model of its surroundings. One reason to use LIDAR rather than radar is that the laser’s high-erenergy, shorter-wavelength laser light better reflects nonmetallic surfaces, such as humans and wooden power poles. Google combines the LIDAR system with vision cameras and algorithmic vision-processing systems to construct and react to a 3-D view of the world through which it is driving (Reference 2).

[rgleason]

Collision Avoidance - waves not cars.
https://www.digikey.com/en/articles/...iving-vehicles

See reccommended miniature radars at the bottom.

Or point your 4G down more. 2 radars?

[rgleason]

Problem with big radar is power required!

[barnakiel]

@ boat alexandra

(This sounds great but is rather vauge on how to actually implement it. I can compute fourier transform of the various inertial measurements, and this does indeed show peaks at various frequencies. These frequencies vary greatly depending on the sea-state, the boat, and the angle it is sailing.)

You cannot (??? my opinion, not a fact) use Fourier as these are not train waves in the understanding of Fourier transform. These are NOT two or more wave trains overlapping. The wave data as seen say at NOAA represents estimated averages, while actual waves in the ocean are individual data points that are distributed around these averages in a quasi normal way. This is to say the majority of actual waves will be anything from 1 to 3 deltas away from the average - whether you take their amplitude (maybe 3 deltas), period (less delta here maybe 1) or direction (delta ?, no idea, probably 10 when cross swell is coming in).

There is a very simple experiment that proves my point: Take real data from any ocean buoy and run the Fourier, then take the same data but a different slot, run Fourier again. Bingo. gigo. Run such a test one day, if just for fun.

If your boat were sailing a virtual ocean rippled by virtual waves which are like compound sound waves (delta 0 in the compounds) then off course Fourier would do and you could make the AP use the data to act accordingly.

(... This type of logic is thrown at me constantly. It is not valid at all. The real reason NKE or whomever didn't do this, is they don't know how or have not put enough effort to succeed. It is not that it can't be done. ...)

Wave data / course optimisation is calculated externally for NKE. I believe MaxSea or Adrena do the calculation for averages (swell, secondary waves, etc) and allows for it in the suggested routing. Then this routing is plugged into NKE and the AP tries to do the rest of the job based on actual behaviour of the boat. NKE has always been at the forefront, just look at how early the implemented g sensors.

Imho: if the core is fast enough to calculate a g model of boat behaviour specific in any time frame and monitor for drift, then g data could be judged against tilt and then possibly maybe a fast and powerful enough arm could preset the rudder for the incoming wave. Just think of how a drone core reacts to windgusts: acceleration + tilt = reaction. Marry this with a real time created wave patter model ...

Yes. I know drones now have 'wave sensors' now (they see the objects around them, or at least ahead of them)

To sum up:

1) I am not 100% sure we need to see the actual waves to react to them. I am convinced seeing just the nearest wave would be very nice though, much as at now we seem limited by our choice of relevant sensors (range! - the wave can be a distance off, + problems with bouncing any wave off a slanted water slope ...)

2) I seriously doubt that Fourier or anything can be applied to de-compose real sea wave conditions (train) into 2 or 3 sub-trains. Simply because such sub-trains do not exist in the realm of compound wave analysis.

I am open to more input on 2) as my knowledge is based on education gained over 30 years ago, before we had PC on every researcher's desk.

I greatly admire your job. And beyond the job I admire your attitude of sharing what you have learned.

[seandepagnier]

As far as using ultrasonic to detect the waves... I think it's a problem. There is too much noise and bad reflections, but maybe it can work sometimes??


For lidar. I think it can work, but where can I actually get a reasonably cheap unit to work with?

[bcn]

Originally Posted by rgleason

Velodyne’s HDL (high-definition LIDAR)-64D laser-sensor system, w

aprox. 80.000$ each
The least expensive Veoldynes are the VLP-16 series and then you talk about 7K$ to start.
The most automotive manufacturers are using them for their autonomous vehicles. We have one of those for mobile mapping.

As resolution (and neither cm precision) is no issue for wave recognition (range might be one) the upcoming linear (not rotating) sensors might be a way.
LiDAR will have to consider mist or dense seaspray.

Hubert

[barnakiel]

Originally Posted by boat_alexandra

(...)

For lidar. I think it can work, but where can I actually get a reasonably cheap unit to work with?

From the car industry. Around 2025 and thereafter.

b.

[barnakiel]

Originally Posted by bcn

(...)

As resolution (and neither cm precision) is no issue for wave recognition (range might be one) the upcoming linear (not rotating) sensors might be a way.
LiDAR will have to consider mist or dense seaspray.

Hubert

Range of these newly developed lidar sensors?

Regards,
b.

[barnakiel]

LIDAR | High End Scanning Lasers & Obstacle Detectors | Obstacle Detection Sensor | Hokuyo Laser Rangefinding - RobotShop

b.

[barnakiel]

Sweep V1 360° Laser Scanner - RobotShop

@ boat alexandra

???

b.

[rgleason]

barnakiel, I was on that site. Frightened by the cost of some of that, but the
Sweep V1 360° Laser Scanner might work, but it would have to be on a gimbal up fairly high, maybe 8-9 feet?
Could you mount just one of them on the backstay to get decent reflections from oncoming waves?

Sean, see the mesh that Lasstools creates from LAS. Would it take too much HP to do that in real time?