Let's talk marine computers
 

I’m in the process of getting a new boat ( http://www.kryptoniteboats.com/image...e/Kryp_465.jpg ) and Kirk needs some specs from me. I’m planning to add a computer. I have space for a 10” or even 14” screen. Computer will be a standalone unit, probably Tactronics ( http://www.tactronics.com ). They are around the corner from where I live. I’ll probably use their displays also. The hardware aspect seems to be under control.

I would like to collect experiences on the type of software that is being used. I mainly use navigational software (with a redundant standalone chartplotter as back-up). My main concern is the user interface to the navigational software. I need the main controls literally at the fingertips of my right hand. For this, I plan to install a marinized numerical keypad plus a small number of dedicated switches. (Trackball and keypad for when the driver is not preoccupied..) The chart software ideally should have an API that allows interfacing with external hardware such as the keypad, or the dedicated switches. If it works with touch screens, that would be great also. I can do the integration work, write software that interrogates the inputs, but I hate to use clumsy keypad macros that may or may not work . I would be grateful for any input.

i don't get it...why??..is this a race boat?? fishing boat???

If you're out running in 5' seas, do really think you'll have time to work a track ball?

I don't think anyone here has an integrated system like you're looking at.
An 8 knot cruiser yes. An 80 MPH Cigarette, no.

I just came from their site.

Geeze!!!!
How much money have you got????

You're looking at ruggedized millitary grade hardware. That stuff is freak'n expensive. I know, I work with similar hardware every day.
I'll bet you'll spend close to half the cost of the boat purchase just in electronics. Not to mention the installation and setup costs. It sure isn't a "do it yourself" deal.

If you plan on long distance running, all you need is a good radar/plotter with a GPS interface. That'll get you any place on this earth. And don't forget, paper charts and basic navigation/plotting skills are still required.

i don't get it either...must be more to the story...i thought i was bad (race boat) with data aquisition systems/gps(s!) and smart craft for the engines.
i want killer electronics for our fishing boat,,,but this stuff....i would rather have triples!

Eyes on the water and gauges.

Left hand steering and K-planes

Right hand throttles.

Feet trying to level the body.

Half the time I can't even hit the next waypoint on my GPS, much less a track ball.

Money better spent on a simple GPS with plotter.

Nice concept though just not for a rippin Go Fast.

Price: It's pricey, but not outrageously. I wish the electronics would cost 1/2 of the boat ... The 10" display is $3K, $4K w/ touchscreen. But it's low power and totally waterproof. Also holds up to 20 G

Trackball: I do NOT want to operate a trackball or mouse in 5ft waves, hence the button interfacing question.

Why: It's not for fishing, nor for any nefarious purposes. I like computers. I can certainly do it myself, I built an ALTAIR in 1975 and many systems ever since. No sweat. I can run a charting program in the foreground and, say, Rinda's Diacom software in the background ...

Aaahhh...........ok! Sounds great to me!

Interesting.

I used to program in assembler clear up through cobalt and c++.

You may be able to get one for free if you do the work with them!

There seem to be a lot of products that offer charting, gps etc at a more economical price.

It would be nice to put a "glass panel" design in a boat to replace some gauges and indicators assuming you have the proper I/O devices, but it would not require much processing.

I'm really a techno junkie but I quite old fashioned when it comes to Go-Fast boats.

It is so hard to run my simple GPS at speed. Normally I just dead recon.

I like the subroutine or scroll through at a touch of a button idea but have it where you can rest a thumb and pinky. I would probably only have buttons at the 2,4,5,6,and 8 position for easy reference. Have fun and send pics.

Baja: Buttons are planned to go right in front of shifter/throttle. "Glass panel" idea is implied in the concept, although the main gauges will be "conventional" Gaffrig SmartCraft gauges. The thing absolutely is not allowed to show any degradation in functionality when the computer goes belly-up.

Sorry, no COBOL wanted :) What I'm looking for is a ready-made application that has a documented programmer-access interface. (Probably too much top ask for ... guess I'll have to hack it by subclassing menus ... nasty, nasty...)

I'll send pictures when it's hatched further.

I guess spell check got COBOL to cobalt.

Have fun, your system should be quite unique.

I'm in touch with http://www.chersoft.co.uk. They offer a COM server that comes close. From the looks of it, it is the basis of certain commercial packages on the market.

More as it develops.

Here's my nickels worth.

While running "hard",
I don't think you will end up needing to do anything other than changing views and info pages:
*Pan right 1/2 screen
*Pan left 1/2 screen
*Pan up 1/2 screen
*Pan down 1/2 screen
*Zoom In 1 step
*Zoom Out 1 step
*Center Screen to current position
*Next page (to scroll thru nav and data pages)
*Back to chart page

All other data entry and functionality should be left to the touchscreen form of entry.

My guess on one way to get close to the above is as follows:
One 4-way joystick paddle for 4-way pan function (a fast double-stroke any direction centers to current position).
One 4-way joystick paddle for Zoom (Up/Dn) and Scroll (Rt/Lt for nxt pg, last pg - a fast double-stroke any direction goes to Chart page).

Nothing as large as a bow-thruster stick, nor as small as a Guest remote spotlight stick. Probably somewhere around an old Bennett single stick tab switch. Could put both of em above your throttles..

Should be easy enough to modify a 8-or-12-button CAD keypad guts to work with the paddle switches. Then a matter of assigning functions to the "buttons" via standard included software for the button pad. (assuming that you can assign hotkeys to the necessary functions in the native charting software you plan to use..).

I like your idea and don't think it sucks.

I DO think you should avoid trying to do anything more than the above-mentioned functionality to the remote pad.

Phd: Your list about covers what I have in mind as the "fingertip subset." The joystick idea is interesting, I'll look into it. As for keyboard macros, I want to avoid them like the plague. Very unreliable. What I have in mind is a daemon that either interrogates dedicated switches (via parallel port, for instance), or it hooks into the keyboard queue. With the proper inputs, it will call the respective functions directly.

We've run 12-and 16-button digitizers for nigh on 20 years in our engineering department. We've found that Virtual Tablet driver/software has performed admirably for us. Some of our engineers only assign button events to F- and ALT- combinations while others assign full macros. You are correct in the macros are questionable, but nobody has ever had any issues with assigning something like (two fast clicks of button #10 returns an ALT-F9 to the system).

You obviously have the I/O translation capabilities to build your own "driver" for the input device. I am still not fully understanding how the charting software will recognize these inputs unless you send them in a hotkey format...

All: I talked more w/ Chersoft), and they say: "We have been involved in developing systems for the RNLI (UK Life Boats). The operators are strapped in and operate the computer using dedicated controls built into the armrest. The design of the hardware is very shock resilient: no fans, no spinning hard-discs, etc. I guess you would set up charts, routes, etc. using a normal mouse & keyboard. Then you would want to put the system into a mode where it displayed the information required by the helmsman with virtually no manual input. The display would need to show the best chart data, the vessel, the route, the next waypoint, distance & CTS, that sort of thing."

That pretty much covers what I had in mind. I'll stay in touch with them.

Here is what I don't like about dedicated chartplotters and how a PC-based system could help.

Too many buttons to fumble with. Keys in front of the throttle would take care of it.

- At high speed and during hard maneuvering, sometimes I lose track of the route. I'd love to have clear steering cues (such as a big yellow arrow) .

- On many chartplotters, redraw rate is anemic. No problem for a fast Pentium.

- The larger the display, the easier to read, especially at high speed.

- A PC is powerful enough to do several things at one. For instance, I plan to monitor and log all parameters of the HP525 that are made available on the CAN-Bus.

And there's always room for other winter projects.... Any requests?


Phd: Hotkeys (as oppsoed to keyboard macros) do work, but the app has to support assignable hotkeys (most use fixed hotkeys .. what the programmer thought to be hot). Also, the app needs to have the focus for keys to work. I'm indeed planning to use a driver (Win 2000 service) that monitors the keyboard and that talks to the app via the COM (Computer Object Model ... not serial port) interface.

I think (hope) I’ve found the solution. It’s the X-keys switch interface from PI Engineering http://www.ymouse.com . This thing translates a switch closure to one or more (programmable) keyboard codes which are then sent to the computer. No software or custom programming required.

I have looked at all the PC compatible navigation packages. Most accept hotkeys. They usually are what the programmer thought are intuitive hotkeys. I found no software packages that allow for definable hotkeys. It would be possible to use a keyboard macro program, or write a dedicated one, but they usually are not reliable.

Writing a custom application, even one that is based on commercial components such as the ones from Chersoft, looks like a bad investment of time if the only problem is the keyboard interface.


With a switch-closure-to-keyboard-code interface, I can map dedicated switches to keys, even key combinations. I can use the same type of switches which are used on my dash (in my case, Contura type), which makes for a sano look.

The PI Engineering units are not marinized of course, will need some work and a dry place. I don’t see major problems there. If the project doesn’t pan out, then I can always use the switch panel for something else.

Also, I won’t have a dorky trackball or a numerical keyboard on my dash. This way, the boat doesn’t look like a cash register.

It’s enough that the boat sometimes feels like a cash register.