Communication protocols, which was what Dave was just asking, so for us, at this stage the predominant one, and you'll see in the schematic in a minute is CANbus, Controller Area Network, so, we trimbles we adhere to J1939 as our standard. There's CANopen out there, which some companies and manufacturers use, which we have to integrate with occasionally. We are getting more to the Ethernet world, the system you see here is four wire. We do have Ethernet, so the main Ethernet bus of this three in the system, two in the system, the main one is from that black box on the floor, which is the Linux, brains of the system talking to the display. Serial, so there are some devices still out there that are serial, so, we have to support that, and then obviously WiFi, is how this system would support a bring your own device. So, if you had a Galaxy Tab, you could actually run our system with a Galaxy, James got one here, So you could run, you don't have to use our display. So, any modern Android device could be used, and you could use a WiFi link, from the Linux box to your display. This is not really designed for people wearing T-shirts obviously, So, lots components, lots cables. This is diagrammatic form showing all the different cables, all the different components, how they hook up, and we actually drew this up for installers, So, you can see all the part numbers of every cable, and then display and stuff for more diagrammatic. Now, you can see here, quite a few things going on, this is specific to an excavator, and you can see really, all the IMUs going down the arm. So, the radios in the cab area, the switches, and display and stuff in the cab, this is the hydraulic side, so the autos and then, we have the GPS receivers next the magnetometer we're talking about, right. out the back, to figure out what the machine is. As low as chance logical charts place to where those passes through a machine. Now, we'll come back to this. It's useful as a reference for when we start. The blue cheat diagram. Yeah, for later. So, we largely use two different communications. So, when you mentioned the IMU, you said it had six axis. Yeah. Well, why don't you use a nine axis meter for more accurate. In terms of [inaudible]. Yeah, so you get integrated nine axis meters, so it has a gyroscope, accelerometer, and magnetometer. The level of calibration and time that we went through when we develop this in this environment [inaudible]. Yeah. Okay. Yeah. So, Trimble and other industries where you're less magnetics or less steel around use magnetometers a lot. We don't much just because there's so much steel and the environment is quite messy, it's quite messy, like electromagnetic interference is quite bad, and stuff like that. So, it doesn't really offer often much. Okay, so we largely within the system right so between the system and the rest of the world, we use WiFi, we use the UHF radios, we use cell that suggests. But, within the system we largely wire things. We do a little bit of wireless, but again the EMI environment is quite nasty, so we normally use quite a lot of wired stuff. There's really two big ones we use which is CAN and Ethernet. It's quite useful to understand that on how much you guys have dug into this, but my big breakdown is really Ethernet, it's more of about, point-to-point and tree types network. You start to see this when you plug things together, and it's more of a bus network has terminators, multiple devices on the same bus, that sort of thing. Have you guys covered CAN in general? Okay. The big thing with CAN, is you have these terminators on the end that define sort of the main bus. Right, and then from that main bus, you have devices coming off the bus, and these are really end up being a two design parameters, of how long the bus can be, so between the two of terminators, and then how long each stub can be, which is how far all fat mainline to each node. When we design, we have basically a bunch of different standards, but we had pretty standard lengths for maximum lengths we can run, and that basically gives us an envelope of how we can run the bus through their network. Our goal often is to try and basically put as many nodes on a bus as we can get away with before we run out of bandwidth. Because, that's shearing the columns lines, it's a lot more components to connect. If you look at all these components, all of these are embedded processors, they all have data to share, they all have measuring stuff. You know if we had to do a point-to-point links for every single one of these, we would have this box in the middle with 50 wires going into it. With sort of a bus network, you can have, the main controller being on the same network as a bunch of sensors, and they can all report it using common wires, which is quite handy. Now, if we start to look at that diagram, and we overlay it with communications networks, you can start to see where the columns goes, and it's small for you guys, but the blue, the green, the yellow, and the red are all CANbuses, so we have up to four CANbuses in the system, and that's just trying to segregate out data because, largely we run out of bandwidth actually. So, you can start seeing the green going all throughout all of these components, and it's sort of picking up data from a bunch of places. You can see the blue going out to the arm, and picking up all these IMUs, and then the orange, and the purple are the Ethemet lines, and you can see these are point to point, from one Linux box to another Linux box, from one down to the display, and they're much more constrained, but they're much, much higher bandwidth. Anybody spot the one funny-looking CANbus on this diagram? [inaudible] So, what Jason is alluding to, is you can see, we have this rainbow pattern going into our main Linux box, and most everything is going there to collect it together. We have this one CANbus down here, which is directly between the GPS receivers. What these guys are doing is they're sending data between themselves in order to very accurately calculate that heading. Now, each one of those cables is a physical cable, which has a bunch of wires inside it, and so, and this is actually, you've taken the stripped down, when it hit the full one actually has more. So, if you actually look at each one of these, each one of these is the individual wires with inside the wiring harnesses, and you can start to trace through the entire system through all the different cables, and how all the components actually are wired together. [inaudible] the system's not working. Yeah, he may have to trace a wire the whole way through to figure out where the fault maybe. Get out a voltmeter, see what's actually connected, and then trace it all the way through, and you'll start to see when you start playing with them, we end up using a lot of connectors on the end here, which actually can be modified, are field serviceable, so they can be pulled apart, repaired, and fixed, or reput on in the field, and that allows the technician to repair it. We caught up and get in situations where you imagine cables going down on the boom of an excavator, and that guy's dealing in water, in mud, in muck, and he's digging in rocks, is just quite a nasty environment, and over time, either he'll damage the cables, or he'll yank it, it'll be yanked too much because he caught on something and swung around and all that just happens remarkable amount.
