Here is an odd question for all you mariners. I am supposed to give some input to students at our university who are building a sailing robot that’s supposed to cross the atlantic this August (http://www.microtransat.org/). One sub-group of the team is tasked with building the collision avoidance module and various avenues are being pursued.:
[ul]
[li]An obvious part of this would be an AIS receiver. This might help avoiding you big guys. Might require some extra brains to also avoid the wake of fishing boats to avoid the fate of one of last year’s competitor to get caught in the nets of a fisherman. [/li][li]If power constrains allow (this is an approximately 4m sailing robot. the collision avoidance people have been granted 2W continuous and 20W max), one could consider an AIS transmitter. So one question is: would you alter your course (slightly) if you saw an AIS target out in the middle of the Atlantic which is an unmanned little robot? Or is it best to assume to be invisible? [/li][li]Lastly, since Radar is out due to power limitations, the students are investigating IR. What are your thoughts on avoiding anything using FLIR? I imagine the image processing would have to be quite sophisticated to figure out what’s in front of the boat. [/li][/ul]
Any input welcome.
There is a book about this called COLREGS 72, prevention of collisions at sea. These are the rules for all vessels and if your unmanned craft fits the description of a vessel, and I think it does, it should also be following the same rules. Perhaps with an AIS transponder you could navigate within the rules but it will only work with other vessels equipped with AIS.
Sorry - I probably didn’t make that too clear: Following the COLREGS is one of the prime objectives in the whole endevaour. The problem arises from detecting a potential collision situation (rule 5) and acting appropriately. The OP is about finding out what you think is the best way to to detect a risk of collision (with a vessel or other objects).
COLREGS
"Rule 5 - Lookout
Every vessel shall at all times maintain a proper look-out by sight and hearing as well as by all available means appropriate in the prevailing circumstances and conditions so as to make a full appraisal of the situation and of the risk of collision."
That International Law conflicts with one of the rules for your Contest:
“The sailboat must be fully autonomous, no operator control is allowed.”
To fulfill COLREGS sight and hearing are both mandatory. I don’t see how you can possibly fulfill both sets of rules.
I agree with the above statement. A completely autonomous vessel could not comply with COLREGS and would thus violate current international law. You can’t keep a proper lookout.
Have you discussed any of this with you local Coast Guard or Maritime Authorities?
[QUOTE=DrSAR;153008]Sorry - I probably didn’t make that too clear: Following the COLREGS is one of the prime objectives in the whole endevaour. The problem arises from detecting a potential collision situation (rule 5) and acting appropriately. The OP is about finding out what you think is the best way to to detect a risk of collision (with a vessel or other objects).[/QUOTE]
I agree with LeakyBrain, autonomous vessels are a disaster unto themselves, and you probably won’t find much sympathy on here because most of us see projects like this as the path to destroying many of our jobs. However, to answer your technical question, the combination of AIS and ARAP (Automatic Radar Plotting Aid) is probably your best bet for collision avoidance. I’m too tired at the moment to go into all of the capabilities of ARPA but basically you can think of it as one big, all-inclusive, collision avoidance computer and damn near every radar manufacturer that exists makes some model of their radar with built-in ARPA.
Speaking of radar, don’t forget to make your boat AS BIG OF A RADAR TARGET AS POSSIBLE through the use of radar reflectors, lots of metal in the construction, lots of built-in right angles, etc. If you play your cards right you can make a row-boat look like a tanker. Knowing what’s coming at you doesn’t mean a damn thing if you think you have right of way but they can’t see you. Having right of way doesn’t mean anything if you look like a small wave on their radar screen and then they still run over you.
simple answer…program your vessel to stop dead in the water anytime there is any kind of target detected and not be allowed to proceed until nothing is detected at all. Should take you about two to three years to get across is my estimate.
-
-
- Updated - - -
-
Here’s a page from their paybook btw…interesting reading especially about our young Navee lads and lass’s from Annyapollis…was their effort funded by the taxpayer, because I didn’t authorize such spending?
About the Microtransat
The Microtransat Challenge is a transatlantic race of fully autonomous sailing boats. The race aims to stimulate the development of autonomous sailing boats through friendly competition.
The 2014 transatlantic race is open to competitors for the duration of the year, please see the registered teams page for a list of teams who are intending to compete.
So far this year two teams have attempted the challenge. A map of both attempts can be seen on the tracking page.
On May 16th the team from the US Naval Academy in Annapolis, Maryland, USA began their attempt from Cape Cod in Massachusetts. On May 22nd their boat was run over by and caught in the nets of a trawler. The crew brought her onboard and she has only sustained minor damage and will be returned to the team. You can read more about their boat on the USNA boat details page.
On September 6th 2014 Team Joker began their attempt from Barton-on-Sea, England. They experienced very light winds and a very strong tide, ultimately ending up being washed ashore on the Isle of Wight during the morning of the 7th of September. You can read more about their boat on the Team Joker boat details page.
On November 30th 2014 Team Joker began a second attempt Boscombe, England. They initially sailed south to their first waypoint, about 12 nautical miles ofshore. They then turned west to sail along the English Channel, towards the start line. At around midnight on the 30th they began travelling backwards and this was believed to be due to a strong tide. Around 6 hours later the boat continued on it westward track. At approximately midday on Decemeber 1st the boat neared the Portland Island and became caught in strong tidal currents. After following two complete tidal cycles on the eastern side of the island, the boat crossed to the south of the island narrowly missing the most southerly point of the island. It was then dragged north by the tide before finally running aground on the western side of Portland Island at approximately 7pm on December 2nd. At around 2am on December 3rd the boat was recovered and found to be mostly intact. For a more detailed account of what happened, see the team’s page.
You can view the route taken by all the boats on the tracking page.
Related CompetitionsA spin off competition, the World Robotic Sailing Championships (WRSC) was created in 2008 and has taken place annually since then. This consists of a series of short races and other challenges for competing boats to undertake autonomously. Many teams who take part in the Microtransat (or who plan to) also attend the WRSC. Also attached to this event is the International Robotic Sailing Conference (IRSC), which allows researchers in the area a chance to meet up and discuss the scientific issues related to robotic sailing.
The SailBot competition is another competition for robotic sailing boats, which has taken place in North America since 2006. This is targeted at student teams and focuses on high performance short distance racing.
Microtransat teams are encouraged to attend a WRSC or SailBot event to meet other teams, exchange ideas and test out their boats and compare against others.
[QUOTE=LI_Domer;153017]I agree with the above statement. A completely autonomous vessel could not comply with COLREGS and would thus violate current international law. You can’t keep a proper lookout.
Have you discussed any of this with you local Coast Guard or Maritime Authorities?[/QUOTE]
Right.
It’s not a joke to put any kind of vessel out into the ocean – you could be endangering lives and property. You could be facing serious liability, including criminal liability. You should read and understand the COLREGS, consult with the Coast Guard, before you do this. In a collision with commercial shipping, it is unlikely that you would do any serious damage, but with a 4 meter drone sailboat you could sink a yacht and kill people. There are even rowboats, from time to time, in the middle of the Atlantic – they have the right to be there (provided they are observing COLREGS themselves) and have the right not to be killed by your science experiment.
I guess that it might be possible to fulfill the Rule 5 requirement of keeping a good watch, but you would need to have some serious telemetry. AIS is not enough. For one thing, not all vessels have AIS. You will need a radar reflector for sure, better an active one (like a “Blipper” or “SeaMe”). You will need radar. You will need some ability to keep a visual watch – I guess a PTZ camera would do it, but do you have the bandwidth to receive the video stream? If you have Fleet Broadband maybe you would.
I suppose collision avoidance could be automated – if you had really good processing of visual, AIS, and radar data (that would be a really big project). Whether the various maritime authorities would accept that as fulfilling the COLREGS, I don’t know. I suppose you could program the vessel to open through hulls and sink if it got into a desperate situation.
Besides the maritime authorities, you will surely want to check with your insurance company.
Sounds like a fun project, but to do this in a legal and responsible way would be quite daunting, if it’s possible at all.
Here’s an interesting paper from NASA on automated collision avoidance: http://www-robotics.jpl.nasa.gov/publications/Yoshiaki_Kuwata/Kuwata11IROS_final_v3.pdf
Which seems to imply that autonomous collision avoidance without human watchkeeping is acceptable. Keep in mind, however, that NASA has no legal authority to interpret COLREGS.
Fishing boats, trawlers, tugs with tows or vessels with broken equipment are always out there and discerning the hazard usually takes a person.
Take a fishing boat that is working with nets. AIS may show the boat is fishing and radar may show its velocity but the nets themselves may be the real hazard. Your computer would have a heck of a time recognizing that. Nets don’t reflect radar, don’t have AIS, are at the surface of the water and are not always behind the vessel or in its wake. Yet a net can be a huge problem.
Then there’s the fact that many smaller craft don’t obey the Rules. Sail boaters and yachts think they own the water and don’t comply. Fishermen will defend their nets and pots by threatening a collision to force you away.
Then there’s things like logs or debris adrift that are just floating trash.
Of course the odds of encountering one of these hazards in a single voyage is slim but should you encounter one it could be disastrous to more than just your science project.
[quote="“DrSAR”"]
Here is an odd question for all you mariners. I am supposed to give some input to students at our university who are building a sailing robot that’s supposed to cross the atlantic this August (http://www.microtransat.org/). One sub-group of the team is tasked with building the collision avoidance module and various avenues are being pursued.: [li]An obvious part of this would be an AIS receiver. This might help avoiding you big guys. Might require some extra brains to also avoid the wake of fishing boats to avoid the fate of one of last year’s competitor to get caught in the nets of a fisherman. []If power constrains allow (this is an approximately 4m sailing robot. the collision avoidance people have been granted 2W continuous and 20W max), one could consider an AIS transmitter. So one question is: would you alter your course (slightly) if you saw an AIS target out in the middle of the Atlantic which is an unmanned little robot? Or is it best to assume to be invisible? []Lastly, since Radar is out due to power limitations, the students are investigating IR. What are your thoughts on avoiding anything using FLIR? I imagine the image processing would have to be quite sophisticated to figure out what’s in front of the boat. Any input welcome.[/li][/quote]
All other things aside, if I saw an AIS target I would be looking for the vessel. This would help especially at night if I could not visually see it or by radar (4m is essentially like looking for a log in the water).
FLIR works well at sea for spotting other vessels. I’ve used it on ships before.
I cannot vouch for fishermen. They are not always “on the look out” since they are focused on fishing much of the time.
Thank you all for the feedback. I appreciate that we should consider the safety of others as the highest priority in this project. And I understand this whole projects appears to be bordering on non-compliance with COLREGS. (we can argue from which direction)
I am hoping that should it come to a collision, the 190kg at relatively low speed won;t be able to do much damage to others. I know that this sounds bad and feeble as a strategy. In the given situation, given that they’ll go anyway, I find it best to advise the team such that their collision avoidance works as well as it possibly could.
The ‘rules’ (http://www.microtransat.org/rules.php) indicate that the race organizers have talked to Irish and US coastguard (they wanted them to put a light on top :-).
WAFI, thanks for the paper. The conclusion that sailing under COLREGS is possible is encouraging but is of course based on the assumption that you see everything. (Their sensory input was RADAR and ‘stereo vision’ - although they are of course moving at higher speed). Will peruse with great interest.
Quimby, I am interested to hear about your FLIR experience. I presume there was a human looking at the data.
One thought about active RADAR reflectors. Does anyone use those to detect vessels in the vicinity? I.e. you get hit by RADAR so someone is out there? Of course that doesn’t cover anyone in stealth mode (radar off, logs, debris, ocean rowing boats) but all that have enough power to run radar continuously would be detectable.
[QUOTE=DrSAR;153098]Thank you all for the feedback. I appreciate that we should consider the safety of others as the highest priority in this project. And I understand this whole projects appears to be bordering on non-compliance with COLREGS. (we can argue from which direction)
I am hoping that should it come to a collision, the 190kg at relatively low speed won;t be able to do much damage to others. I know that this sounds bad and feeble as a strategy. In the given situation, given that they’ll go anyway, I find it best to advise the team such that their collision avoidance works as well as it possibly could.
.[/QUOTE]
Yeah… cold comfort for the fishing guys whose gear you just fouled… or someone you sank because more damage occured than you thought would happen… ooops. I say you have a manned vessel escort the unmanned vessel… you could step in and intervene if there were a hazardous situation. Only way I see for you to truly make sure you don’t cause harm.
Like a self driving car… I don’t think google just put their car in downtown manhattan for it’s first trial version… you need extensive testing imho
you know, this should not even be a discussion because anything out there without manning aboard to control it is a derelict or nothing other than debris and this entire project is rather ridiculous folly by a bunch of nerds trying to prove something without true value to the maritime world.
personally, I hope every one these contraptions that is sent out on its perilous voyage ends up being run down if not accidentally, then deliberately!
note to the original poster…why don’t you try to put your energies into useful applications of technology at sea instead of this nonsense?
personally, I hope every one these contraptions that is sent out on its perilous voyage ends up being run down if not accidentally, then deliberately!
Hrm. This is a slightly dark sentiment. I’m glad you won’t be in that particular stretch of water. (And maybe we should ask the race organizers to delay publishing position reports).
I don’t quite follow that this is ‘without true value to the maritime world’. After all it’s those nerds who came up with technology that you now use when out there. In this particular case, these kids are learning skills that will hopefully help them develop useful marine technology in the future. I also don’t buy into the Luddite attitude that feels threatened by these ‘machines’ replacing you at the wheel. It’s people who come up with manning rules etc and it is probably better to lean on organizations that write those rules.
Coming back to the original question, I take away that the team should throw all that’s reasonably within the energy budget (AIS, vis, FLIR, RADAR and high (RADAR) visibility) at this problem. Thanks.
[QUOTE=DrSAR;153103]Hrm. This is a slightly dark sentiment. I’m glad you won’t be in that particular stretch of water. (And maybe we should ask the race organizers to delay publishing position reports).[/QUOTE]
Dark sentiment indeed but you are obviously new here and don’t know how I earn my living…
I do not care one bit if your vessels are somehow being controlled by a computer with umpteen different sensors…the COLREGS have absolutely zero provision for an unmanned vessel and DO NOT NEED TO BE REWRITTEN! Anything floating on the surface without a human on it is debris or derelict and no rule whatsoever will apply to them attempting to avoid collisions on their onboard computers. Yes, I am a Luddite and do not want the seas to become filled with drone vessels like the air has become…pardon me all to HELL for my opinion!
as far as these youngsters honing their technological talents with this projects is a hollow argument. There are many hundreds of other projects they can be engaged in than this. How about health related technology where people’s lives are saved or extended? How about water purification? How about ways to protect endangered wildlife in Africa and Asia from poachers? Hmmm?
That’s all we need are more Mister Hoopers with their cityboy hands and dreams of being the next wunderkind billionaires!
[QUOTE=c.captain;153104]Dark sentiment indeed but you are obviously new here and don’t know how I earn my living…
I do not care one bit if your vessels are somehow being controlled by a computer with umpteen different sensors…the COLREGS have absolutely zero provision for an unmanned vessel and DO NOT NEED TO BE REWRITTEN! Anything floating on the surface without a human on it is debris or derelict and no rule whatsoever will apply to them attempting to avoid collisions on their onboard computers. Yes, I am a Luddite and do not want the seas to become filled with drone vessels like the air has become…pardon me all to HELL for my opinion!
as far as these youngsters honing their technological talents with this projects is a hollow argument. There are many hundreds of other projects they can be engaged in than this. How about health related technology where people’s lives are saved or extended? How about water purification? How about ways to protect endangered wildlife in Africa and Asia from poachers? Hmmm?
That’s all we need are more Mister Hoopers with their cityboy hands and dreams of being the next wunderkind billionaires![/QUOTE]
CCaptain you should go out there and see an unmanned boat and claim salvage on it
[QUOTE=c.captain;153104]Dark sentiment indeed but you are obviously new here and don’t know how I earn my living…
I do not care one bit if your vessels are somehow being controlled by a computer with umpteen different sensors…the COLREGS have absolutely zero provision for an unmanned vessel and DO NOT NEED TO BE REWRITTEN! Anything floating on the surface without a human on it is debris or derelict and no rule whatsoever will apply to them attempting to avoid collisions on their onboard computers. Yes, I am a Luddite and do not want the seas to become filled with drone vessels like the air has become…pardon me all to HELL for my opinion!
as far as these youngsters honing their technological talents with this projects is a hollow argument. There are many hundreds of other projects they can be engaged in than this. How about health related technology where people’s lives are saved or extended? How about water purification? How about ways to protect endangered wildlife in Africa and Asia from poachers? Hmmm?
That’s all we need are more Mister Hoopers with their cityboy hands and dreams of being the next wunderkind billionaires![/QUOTE]
Eh, it’s a neat project. Not everything has to save the world, and I can see this developing into beneficial bridge software, or improved gear for things like spacecraft/used boosters, lifeboats, buoys, shipping containers, or anything else you want to know about but can’t spend a lot of space, power, or money on. ECDIS didn’t empty the bridge and AutoChief didn’t clear the engine room. I’m on the lookout, but I’m not panicking yet.
I would be interested in how the lawsuit would look if it was recovered as salvage despite being completely intact, though!
I have been involved in the commercial side of shipping for the past 25 years, prior to that I did a circumnavigation on a plastic…errr, fiberglass ketch. I ran an oceanographic research vessel that tended manned submersibles for many years. Your ASV (Autonomous Surface Vessel) is somewhat akin to a research submersible or AUV on the surface. Limited reaction capability and limited mobility, and an inability to comply with COLREGS. We always had a mother ship there to hover over the manned or unmanned submersible to prevent it’s coming to harm before getting it safely back on deck or under tow.
To answer your early question, if you equipped your ASV with an AIS transceiver named something like “Unmanned Experimental Vessel” it would not be struck by any commercial vessel equipped with AIS. (unless C-Captain was at the helm ;)) At the specified size and wind powered it is effectively stationary to most commercial vessels. It is also extremely unlikely to cause any significant damage to a commercial steel-hulled vessel.
However, this object would be a major threat to just about any fiberglass or CF hulled vessel in the world, including just about every cruising or racing sailing vessel. Most of these vessels are functioning under significant power constraints (like your ASV) with no active electronics (AIS or RADAR) and often skimpily manned. In addition to being low to the water, helmsmen on such vessels have limited visibility in the best of times. I can easily envision a fast cruising sailboat hitting your ASV in the middle of the night on a high-speed reach and being holed and rapidly sunk with potential loss of life.
Even if you light it up like a Christmas tree and equip it with every navigational do-dad known to man, I don’t see how you could reduce that risk to an acceptable (acceptable to me anyway) level.
My advice would be to give the teams a free reign, and find someone willing to play escort to the experiment with a yacht or other vessel, either institutionally or privately.
I think there is someone on this board with some small research vessels available for charter at reasonable rates. 
Feel free to message me with any questions, I will try to get back to you when I next have internet.
DrSAR, just an fyi, the folks on the gCaptain forum, especially c.captain, don’t sugar-coat anything. Be prepared for a very honest response no matter what it is.
Personal or vulgar attacks are forbidden however and we take those seriously.
That said…
I have a friend who was involved with SailBot at the U.S. Naval Academy, but I never really spoke to him too much about it. What’s the overall goal of this project?
[QUOTE=Slowsailor;153112]I think there is someone on this board with some small research vessels available for charter at reasonable rates.[/QUOTE]
Well since I lost that lucrative charter with Shell to be their great white shark monitoring vessel in the Chukchi Sea, I have been thinking I’d see what other work I can find for the mighty ORCA. How does $25k/day sound to you plus you have to pick up the mob/demob back to Puget Sound and all the fuel, the grub plus drinks all around for everyone in a divey bar at some oddball port.