can a robot "hack" another robot

ReVeRsEReVeRsE Posts: 6
edited 2004-11-29 - 17:14:01 in Robotics
········ Can a robot realy "hack" another robot? That's the question I have in my mind since I discovered that robots can comunicate through IR sensors.
I am aware of the techniques used to disconect a client from IRC, for example (It doesn't mean I do it). Is it possoble for a robot to dislabe the IR sensors of another robot (for example·a BOE-BOT). But, there is one thing that must be present for such thing: a PING.
·········So my question is: do Robots that comunicate use PING? If not?·Well,·the creation of a·program that·let a·robot hack another·robot will maybe be hard. So? Do robots use PING? Or a thing that looks like·a PING·and makes the robots receive a reply from the PING?
········ Thanks for answering. smile.gif


  • allanlane5allanlane5 Posts: 3,815
    edited 2004-11-23 - 01:17:49
    A robot (let's call him Rob-A) can only hack another robot (let's call him Rob-B) IF and only IF Rob-B has the 'hooks' for being reprogrammed on the fly, and Rob-A can get access to those hooks. Somebody has to program in those hooks.

    With the BS2, the programming port is the 'Port 16' SERIN/SEROUT port. It requires a physical connection. There is no other programming port.

    Concievably, you could program Rob-B to accept a configuration update from a remote using IR (what I call a 'hook'). If you did this, then Rob-A could change Rob-B's configuration on the fly. It would not be a smart thing to program Rob-B this way.

    The short answer is: no. Robot's using IR to communicate typically use some proprietary protocol that only the two robot's know about. And it would be a really dumb thing to *allow* one robot to reprogram another, especially competition robots.

    "Ping" is a TCP/IP protocol packet. Most simple robots do not do TCP/IP, because they don't have to. So they do not use 'Ping'.
  • ReVeRsEReVeRsE Posts: 6
    edited 2004-11-23 - 14:15:47
    Thanks for your reply allanlane5. I can understand a litle bit more of the comunications of robots.

    I have another example for a robot to disturb another robot, but I need to know some things that I can't find on the net: 1- do robots use some sort of IP? Some sort of ID?
    It is very important to a program that I'm righting.
    2- When a robot uses ultrasonic range devices, the robot send a signal then it receives the reply of this signal. Am I wrong? If yes, robots must have some sort of IP otherwise other robots would receive the signal that another robot sent. No?

    Thanks for answering.
  • BeanBean Posts: 8,076
    edited 2004-11-23 - 15:27:27
    I'm not sure what your getting at ?
    Just come right out and ask what you want to do.

    Sounds like you want to "jam" the other robot ?

    That depends on how well the other robot's hardware is "jam resistant".

    You could attempt to jam the IR communication by just sending out a constant strong IR signal, same with ultrasonic, just constantly send out ultrasonic sound.

  • danieldaniel Posts: 231
    edited 2004-11-23 - 15:53:32
    If the intent is not to just·"jam" the other robot·but implement some sort of "stealth"·cloak for the first robot or to just "confuse" the other robot, it would be theoritically possible to implement reactive and/or proactive IR, sonar, etc. outputs to feed the other robot's sensors with mis-information.· Although either type of goal would require knowledge specific to each of the other robots to be affected, doing cloaking on the first robot would be the more difficult of the two, given the typical sensors and processing power of current hobby-level robots.

    Of course, this would, like in the real world, just lead to an escalating spiral of·sensing technologies being pursued by evasion technologies.

    I, too, am curious, as to your application requirements.

  • allanlane5allanlane5 Posts: 3,815
    edited 2004-11-23 - 17:03:21
    If your "IP" means "Internet Protocol", then no, most robots don't use IP. Ultrasound *does* send a 'ping' -- but it is a submarine sonar-type ping *noise*, not a TCP/IP 'ping' packet. The robot ultrasound hardware listens for the echo of the 'ping' to come back (having physically bounced off some object), and translates that to a distance to an object -- which is *exactly* how sonar does it.

    A TCP/IP 'ping' packet does require a reciever on the 'far end' to recieve the packet and issue a response. With a robot ultrasound 'ping' noise, it does not require a reciever, just an object the noise will 'bounce' off of.

    If Rob-B is using ultrasound to detect objects, and Rob-A has the same ultrasound hardware, then Rob-A can 'confuse' Rob-B's distance measurement by sending a 'ping' when Rob-B is listening for his own echo. Most ultrasound hardware has its own processor to find distance. Hacking this hardware so it *first* listen's for another robots 'ping', *then* sends out its own signal to confuse the other robot, is usually beyond what the original vendor intended. This means it may not be trivial, and it may not be possible at all.

    In any event, you'd be telling Rob-B that you are *there*, just *maybe* confusing it as to how far away you are. This is of limited value. Just as car radar-gun jammers don't make your car invisible, just confuse the radar-gun as to how fast you are going (car radar-jammers are illegal in most states, and it is *really* obvious to the operator when you have them on).
  • ReVeRsEReVeRsE Posts: 6
    edited 2004-11-24 - 12:43:31
    Well, I'm then going to explain a litle better the program I have in mind
    This afternoon, I will develope the program into a formal language.

    See you all.
  • ReVeRsEReVeRsE Posts: 6
    edited 2004-11-27 - 14:09:51
    the program would be like this:

    the robot-1 restart his frequency to minimum and starts ading 1 to it, after it would send a PING to all these frequency Identic to his. If he would get a reply, he would do fase 2, otherwise he would repeat fase 1.
  • ReVeRsEReVeRsE Posts: 6
    edited 2004-11-27 - 14:16:47
    in fase 1, the robot's frequency is a constant. (in previous message, change the words 'frequency' for 'IP')

    fase-2: so the reobot received a ping reply from the same IP he has. Now he's going to send a big quantity of PING's to the IP he has, then he changes his IP to his original IP he had. I think that after this, the PING reply that our robot had sent would wana return to the IP that has send the PING, so he goes to the frequency of the other robot with the IP that robot-1 had. I think it should do the same thing that it does on a IRC channel. No?
  • allanlane5allanlane5 Posts: 3,815
    edited 2004-11-29 - 17:14:01
    It is still not clear what you are trying to do. You are talking in 'solution' terms, when your problem is still not defined. For instance, your problem could be "I want to control one robot from another". Your problem could be "I want to *find* one robot from another".

    Your references to an 'IRC' channel seem to imply you want two robots to 'talk' to each other, for some undefined purpose. It might be helpful if you start talking in terms of 'bytes' of data transferred, instead of IRC or IP PING packets.

    IRC and IP PING packets or port numbers imply a *much* higher level of software than is typically included in robots. Also, solving any of the above problems don't require that much software.
Sign In or Register to comment.