Automatic targeting systems

Discussion in 'Research and Development' started by JustinScott, Dec 23, 2006.

  1. JustinScott

    JustinScott Well-Known Member

    Joined:
    Oct 20, 2006
    Posts:
    2,203
    Location:
    Dallas
    I've seen plenty of people discussing the idea of automated targeting systems... a recent proposal suggested using lasers:

    ----

    Speaking as someone who has done this type of work before. (I built a license plate detection system back in college) What you are discussing is possible and the approach you've taken is on the right path... You've taken nature and constrained it into something that can easily (less difficult?) be analyzed by a system that doesn't 'understand' what a 'boat' is.

    You have to remember how much processing power is required... A couple 18F series PIC working in parallel is required just to do 'simple' digital audio bandpass processing. You would probably end up having a windows box sitting on shore & your boat would just be transmitting the signal for processing. Which would work out in the end since all commands are coming from shore anyway. A PC would also make debugging easier...

    Your luminescence idea is on the right path, but you might be better off searching the video input for the correct 'wavelength' (color) of light emanating from your laser's reflection. Infact, You should be able to find a sensor (not video) capable of detecting only that wavelength. And use the intensity & direction for target bearing and distance. I've use a similar sensor to detect 'candlelight' (& only candlelight) when I made a 'firefighting robot'. If it found signal, it returned an analog (0-5V) depending on strength.

    If you found such a sensor, the logic would be simple. If voltage is above a certain 'X' threshold you have successfully targeted a ship within range of your guns. It would be up to the captain to determine if it was friend or foe. You would use the amount above that voltage to determine the distance to target.

    We have to consider the boat won't be floating level as its fighting... So you would need some sort of auto leveler...

    The target ship's roll wouldn't really be an issue... It won't be changing its distance that drastically from roll.

    I liked the idea of mounting a camera straight up & only putting a mirror in the fray... Auto leveling a single mirror would be easier than quickly adjusting a bunch of electronics & since we are talking about light, performance shouldn't be effected... Infact, by changing the shape of the mirror; you could improve accuracy.

    The range calculations would have to be done by a microprocessor... So you have some processing power. And you have you ships roll in the direction to target from the mirror adjustments... So using the current roll of the ship, the distance to target, and the characteristics (fall rate) of your projectiles, you should be able to calculate the gun's barrel elevation just before it fires.

    Its not clear to me if we could get away with one mirror or if each gun would have to be independent. Since there is some slack in what we are trying to do, I would think we could 'fudge' a single mirror to work. But then again we are just being theoretical... (I think we will start seeing such systems 'popping up' in a few years.)
     
  2. JustinScott

    JustinScott Well-Known Member

    Joined:
    Oct 20, 2006
    Posts:
    2,203
    Location:
    Dallas
    What other ideas have people come up with?
     
  3. aroeske

    aroeske Member

    Joined:
    Nov 24, 2006
    Posts:
    32
    Personally I am against any automatic targeting system. Takes all the fun out of it in my opinion.

    Andy
     
  4. JohnmCA72

    JohnmCA72 Member

    Joined:
    Nov 10, 2006
    Posts:
    681
    I think it shifts the fun, away from the "do-ers" & "fighters" toward the "builders" & "tinkerers". Different strokes, I guess. They seem to be pretty quiet, for the most part, but I believe that there has to be a sizeable group that's interested in the game-play aspects of the sport.

    People need to give thoughtful consideration, before they make massive "improvements", to what the effect is going to be on the game, & especially on all the players who DON'T go after the "cutting edge" of technology.

    For me, the biggest downside is that this sort of thing seems to promote a very high "talk-to-action ratio", which I find personally distasteful. I like getting out on the water & trying to impose my will on others through combat! The more emphasis on "advanced technology", the closer the game gets to being something virtual rather than actual. When it gets to a certain point, it might as well be a video game & we can all just stay home & "fight" from our computers.

    JM
     
  5. iamscottym

    iamscottym Member

    Joined:
    Jan 19, 2007
    Posts:
    39
    The system I'm working on uses a camera on a pan/tilt mount on top of the super. Using the assumption that the pond is flat, and knowing the height of the camera (gun director) from the water as well as it's angles of rotation and elevation you can calculate where it's point on the pond. Then by knowing the height of the guns off the water, their horizontal distances from the camera, you can calculate what bearing/elevation to assign to the guns to hit the target.

    In the future I will add an accelerometer used as a 2d tilt meter to compensate for the ship's listing.

    I will be using a laptop for the video feed and control interface, via a data radio. This will allow me to manually adjust the heights of the guns off the water as the ship takes on water.

    At the moment I'm using an IFI system, but I will probably move to a BasicX once I get a ship built.

    I had considered an ultrasonic rangefinder like the devantech srf08 or a true laser rangefinder, but found them to be unworkable solutions. US rangefinders do weird things when pinging off water, and a laser rangefinder is too big and too expensive to use.

    I currently do not have a ship, and am developing my fire control system by using cheap pen lasers in place of guns. I have plans for a 1:72 Iowa class, but am considering going with a Montana class because I have found the raised turrets provide more depression. (yes, I realize how big these ships will be, but I find the 1:144 clubs near me to be too restrictive). I'm an hour NW of Chicago, btw.
     
  6. JustinScott

    JustinScott Well-Known Member

    Joined:
    Oct 20, 2006
    Posts:
    2,203
    Location:
    Dallas
    Iamscottym,

    What does your ID mean? :)

    Sounds like you have a bunch of sensors, but I'm a little unclear how your are going to use the information coming it. Surely you haven't found a way to decode a video stream using a basicx? IE, how do you know the "horizontal distance from the camera"?

    Chicago, your best bet is probably the MBG group in indiana; which is 1:144 big gun. Or is that the group that is too restrictive?
     
  7. iamscottym

    iamscottym Member

    Joined:
    Jan 19, 2007
    Posts:
    39
    JustinScott

    I'm not using the camera as some sort of 'range finder' like you were talking about. The way I'm finding the distance to a target is actually quite simple. The key here is that we are always shooting at a target that is in the level plane of the pond. You have the camera mounted somewhere on the ship. By using a 3d coordinate system, we call the camera's position (x,y,z)=(0,0,H) where z=0 is the plane of the water. Then, we call the angles of rotation and elevation R and E. Now, using simple trig I can determine the location of the target on the 2d plane of the water. Then, by knowing the locations of the guns relative to the camera (ie: using 3d coordinates), we can calculate the appropriate angles of rotation and elevation to fire on the target. I'm working on an excel spreadsheet with the formulas and a diagram to explain all this. When I finish, I'll post it here.

    MBG doesn't allow any kind of targeting system, so I don't want to join it. There's a new club out here called the Mississippi Battle Squadron, it's basically a few QO guys that moved out here. They also compete in QO ocassionally. Building the ship will be a pain, but after that it's not that much trouble. Plus, I'll have obscene amounts of room for components.
     
  8. Robert Clarke

    Robert Clarke Member

    Joined:
    Jan 3, 2007
    Posts:
    42
    Automatic targeting systems are illegal under IR/CWCC rules.
    However, if I was going to do it, I'd just get a laser pointer and mount it so that it aligned with the cannon barrel. When I saw a little red dot on the side of the hull, I'd shoot. (Assuming it was visible in daylight)

    Not quite so fancy, but cheap.
     
  9. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522
    How about, rather than an automatic targeting system, an automatic gun director system? We've discussed this before on the Big Guns yahoo group, and the basic idea is this: the human operator eyeballs the range and bearing from his ship to the target, and enters that into the transmitter. Then the ship, knowing where the human thinks the target is, brings all guns that can bear to point at the indicated point. For example, say the human thinks the target is about 45 degrees off the starboard bow, about 1 ship length away. The human enters that data (most common idea was a rotating dial for bearing, and a slider for range) into the transmitter. The ship (say an Iowa) receives this data, recognizes that it can bring both bow turrets and the stern turret to bear on the target, then aims each individual turret at that point. The human says "that looks about right", presses the trigger, and unleashes a full nine-gun blast. A bit later, the target crosses the bow of the attacking ship, so the new range and bearing is 10 degrees off the starboard bow, about 1/2 a shiplength away. The human enters the new data, and the ship adjusts its aim. However, since it can no longer bring the stern turret to bear, that turret automatically returns to its rest position, facing directly off the stern. The two forward guns aim, the captain pulls the trigger, and only the bow guns fire. Later, the target is 45 degrees off the port bow and the stern turret can come to bear again, so it rotates to resume firing.

    The current system (where bow and stern turrets rotate together, stop 90 degrees off centerline, and fire separately)is not at too much of a disadvantage against a director-based system because both depend on the Mk I model O eyeball for targeting data. For both systems, the skill of the captain is still the primary determining factor in accuracy, and all the second system does is bring more barrels to bear. In either case, a misjudgement in range or bearing will guarantee a miss. The Gun Director system is generally considered more fair than automatic targeting systems for this reason, as well. Lastly, the Gun Director system must be developed before any automatic targeting system, because once your automatic targeting system discovers the range and bearing to the target, you need some way of getting that data to the guns and the Gun Director system does exactly that.
     
  10. JustinScott

    JustinScott Well-Known Member

    Joined:
    Oct 20, 2006
    Posts:
    2,203
    Location:
    Dallas
    I agree with Carl. The gun director would have to be developed before any auto tracking system is created. And honestly it is a far better solution to the "automatic" systems that I've seen presented. However, I've seen plenty of talk about this system; but nothing concrete. I have a guess that the "secret" radio that NTXBG is creating will be based on this concept.

    Futher, John (from NTXBG) created a paper on this concept. I can probably find the link if anyone is interested. I agree with what he wrote with the exception that I think a system like this should only be able to track 1 opponent at any 1 time. I have long dissagreed with the idea of tracking multiple targets; it is too confusing & too many things to keep track of that will wind up confounding you at a critical moment.

    The last thing you want is trying to fight while your fingers are flying all over a keyboard.
     
  11. happy

    happy New Member

    Joined:
    Mar 9, 2007
    Posts:
    9
    sounds to complicated, just mounta camera on a guna nd zero it then draw the crosshairs on the monitor, simple and fum to use
    lol, jordan
     
  12. JohnmCA72

    JohnmCA72 Member

    Joined:
    Nov 10, 2006
    Posts:
    681

    Somebody has a shore battery that's aimed just that way. However, they make the guy sit in a little shack, so that he can't see anything EXCEPT his video.

    I think that's a fair way to implement such a system on a ship as well. Give the captain a choice: He can have video "down-the-barrel" sighting, or he can have the broader tactical picture - he can't have both.

    JM
     
  13. happy

    happy New Member

    Joined:
    Mar 9, 2007
    Posts:
    9
    lol, that would suck
     
  14. happy

    happy New Member

    Joined:
    Mar 9, 2007
    Posts:
    9
    i'd put a pan camera on the superstructure so i could see everythining, but you would feel isolated alone in a shack and the best thing about the hobby is making new freinds and you can't do that in a shack!
     
  15. JohnmCA72

    JohnmCA72 Member

    Joined:
    Nov 10, 2006
    Posts:
    681

    I think they let him come out between battles, or others go in to keep him company. The idea is just to deprive him of the advantage of being able to see exactly what he is shooting at AND the overall tactical picture at the same time.

    JM
     
  16. happy

    happy New Member

    Joined:
    Mar 9, 2007
    Posts:
    9
    it wouldn't be that bad of a tactil disadvantige but on a ship people could sneak ip on you and you wouldnt see!
     
  17. JohnmCA72

    JohnmCA72 Member

    Joined:
    Nov 10, 2006
    Posts:
    681

    Yep. All sorts of things could be going on that you couldn't see. There might be a fat, juicy target just outside your field of vision. One of your buddies might need help & you can't find him in time. Then there are those enemies who understand your system's weakness & how to exploit them!

    But if you're going to take a special advantage, like being able to get a positive, direct line of sight down your guns to your target, you should expect to give up something in return.

    JM
     
  18. happy

    happy New Member

    Joined:
    Mar 9, 2007
    Posts:
    9
    i don'y agrre in world war to the had variuse kinds of targeting sytems,
     
  19. Tugboat

    Tugboat Facilitator RCWC Staff Admiral (Supporter)

    Joined:
    Jan 10, 2007
    Posts:
    8,298
    Location:
    Statesboro, GA
    That may be true, happy, but in our hobby, you'd have a hard time convincing everyone else to give you a special advantage without giving up something in return...
     
  20. pew-pew-pew

    pew-pew-pew Member

    Joined:
    Mar 18, 2007
    Posts:
    286
    very true... it would be kinda weird to show up with a laser targeting system with out giving something up... at least two battle units