Slashdot Mirror


Your Next Car's Electronics Will Likely Be Connected By Ethernet

Lucas123 writes "As the sophistication of automotive electronics advances, from autonomous driving capabilities to three-dimensional cameras, the industry is in need of greater bandwidth to connect devices to a car's head unit. Enter Ethernet. Industry standards groups are working to make 100Mbps and 1Gbps Ethernet de facto standards within the industry. Currently, there are as many as nine proprietary auto networking specifications, including LIN, CAN/CAN-FD, MOST and FlexRay. FlexRay, for example, has a 10Mbps transmission rate. Making Ethernet the standard in the automotive industry could also open avenues for new apps. For example, imagine a driver getting turn-by-turn navigation while a front-seat passenger streams music from the Internet, and each back-seat passenger watches streaming videos on separate displays." This might get us into trouble when the Cylons show up.

180 comments

  1. Huh? by kuhnto · · Score: 1

    What is it... this ethernet you speak of?

    --
    "A 'person' is smart. 'People' are dumb, panicky animals and you know that."
    1. Re:Huh? by Anonymous Coward · · Score: 0

      802.11 Wireless Ethernet is still Ethernet. Wireshark confirms it.

    2. Re:Huh? by ISoldat53 · · Score: 1

      10Base5

    3. Re:Huh? by epyT-R · · Score: 3, Informative

      Connectivity used when shit has to just work all the time, regardless how many hipsters are in the area.

    4. Re:Huh? by sunderland56 · · Score: 4, Insightful

      Ethernet may work all the time - but there are no guarantees on packet latency. The basis of ethernet is that all traffic is equal; nobody has priority.

      Which, to me, sounds all wrong. I'd much rather the packet from the collision-avoidance system to the brake system saying "holy shit stop NOW" gets higher priority than the next packet of Justin Bieber headed to the back seat.

    5. Re:Huh? by davester666 · · Score: 3, Funny

      Whatever you do, do NOT let the token fall out of the network, because you'll never find it in all the crap on the side of our roads.

      --
      Sleep your way to a whiter smile...date a dentist!
    6. Re:Huh? by gewalker · · Score: 2

      Phhtt, the packet to turn off Justin Bieber should have the highest priority, forget the collision avoid avoidance system, the Bieber avoidance system is more important. You do not want to have to explain to God that, yeah, in my dying moment I was listening to Bieber.

    7. Re:Huh? by Opportunist · · Score: 1

      I wouldn't mind that. He'd probably decide I suffered enough on earth already and I can bypass purgatory.

      --
      We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
    8. Re:Huh? by kimvette · · Score: 2

      Who says it has to be the same LAN?

      Also: CAN can already be interrupted by infotainment systems, since some vehicles have actually used the head unit (the part of the stereo you control) as the CAN bus hub. Idiotic design.

      --
      The Christian Right is Neither (Christian nor right). See: Matthew 23, Matthew 25, Ezekiel 16:48-50
    9. Re: Huh? by Anonymous Coward · · Score: 0

      The 1990s called: They'd like to tell you about IEEE802.1p.

    10. Re:Huh? by Anonymous Coward · · Score: 0

      That's old info: Ethernet has supported 8-levels of priority since the late 90s. Not only that, many little embedded switches (the sort you solder onto a PCB, not the ones you cable up in a rack) support both the priority field in a VLAN tag, and some even look at the IP TOS/PREC (DSCP), and they use these fields to both buffer and schedule frames. Ethernet is bidirectional now (no collisions), and the latest standards even support per-priority pause (although that hasn't made it onto low-cost discrete switch chips. It'll run on copper or fiber (or radio), at speeds from 10Mbps to 40Gbps (with 100Gbps in the lab).

    11. Re:Huh? by inasity_rules · · Score: 1

      This is why, unless the manufacturers are really stupid, they'll use ethercat, which does have guaranteed latency and priorities. But car systems will only use a fraction of the bandwidth.

      --
      I have determined that my sig is indeterminate.
    12. Re:Huh? by Z00L00K · · Score: 1

      Which means that you use separate segments.

      B.t.w. Bieber sucks worse than Beta.

      --
      If builders built buildings the way programmers wrote programs, then the first woodpecker would destroy civilization.
    13. Re:Huh? by hvdh · · Score: 1

      Except for the one master, all slave devices on the Ethercat bus need special Ethercat controllers.
      You cannot hook up regular Ethernet controllers there, and hooking up a modern
      car media player/navigation/controls embedded PC with Ethernet won't work.
      I think that's one of the goals when moving to Ethernet.

    14. Re:Huh? by inasity_rules · · Score: 1

      I think the point is that a standard network is generally the wrong thing for control - therefore you would not want to hook up standard car navigation/media player devices to your control network, but non standard ones which "get out of the way" when necessary. Ethernet (despite all it's nice features) is really unsuitable for control. I know - I've played around with remote IO on ethernet. Devicenet (aka CANBUS for industry) beats the pants off it in terms of latency at lower baud rates on all but dedicated networks. Wrong tool for the job really.

      --
      I have determined that my sig is indeterminate.
    15. Re:Huh? by Anonymous Coward · · Score: 0

      Ethernet (it's a proper noun, bitch) defines layer 1 and layer 2 stuff. Routing and QoS are an entirely different topic, dimwit.

    16. Re:Huh? by kuhnto · · Score: 1

      I was being sarcastic...

      --
      "A 'person' is smart. 'People' are dumb, panicky animals and you know that."
  2. Imagine by Anonymous Coward · · Score: 4, Insightful

    "For example, imagine a driver getting turn-by-turn navigation while a front-seat passenger streams music from the Internet, and each back-seat passenger watches streaming videos on separate displays."

    Imagine!

    Except they're already doing it now on their fondleslabs.

    1. Re:Imagine by bob_super · · Score: 4, Insightful

      Yes, it's a stupid summary, probably from someone who doesn't have a clue on what the current buses do.
      Nobody's saying "Man, I wish my CAN bus had more bandwidth so I could stream!

      And really, people, if you're going to change the bus, can't you make the new one based on plastic fiber and cheap LEDs, so that we stop having fried computers every time a cable gets bad?

    2. Re:Imagine by Mashiki · · Score: 5, Interesting

      What's messed up is the article itself, especially in the belief that automakers will want to switch to this. Right now Cadillac and Lincoln cars have been using fibre in their cars for the 'drive-by-wire' system for years. As well as in parts of the HUD, and rear-display systems. Beleive you and me, they want to use this, because it's reallllly expensive it if gets toasted, and they have to replace part of the harness. This isn't really a job your layman can do, compared to say pulling and restringing an entire wiring harness inside the cab. That's something anyone with a bit of patience and weekend or two can do.

      --
      Om, nomnomnom...
    3. Re:Imagine by ThatAblaze · · Score: 3, Funny

      What a fantastic idea! By switching to a well understood standard it will finally make our vehicles trivial to hack!

      Ohh.. wait.. maybe that isn't such a good thing? *shrug*

      Well I look forward to the day in which I no longer have to call a taxi, I can just take out my laptop and make one drive to my location.

    4. Re:Imagine by benjfowler · · Score: 3, Insightful

      If somebody determined to get up to mischief, they'll do it anyway, regardless of whether or not your car is networked over CAN or Ethernet.

      What you're suggesting, is that security by obscurity works, not that CAN is all that obscure...

    5. Re:Imagine by Ralph+Wiggam · · Score: 1

      Every ethernet network is trivial to hack?

    6. Re:Imagine by jrumney · · Score: 2

      Ethernet is a well understood standard? I think very few people have had to deal with transport level details of ethernet since the advent of ethernet controller ICs in the 1980's (or maybe early 1990's). Most hackers work at higher levels which wouldn't necessarily be deployed on a network that needs guaranteed delivery timing.

    7. Re:Imagine by ThatAblaze · · Score: 1

      I would like to introduce you to my "friends."

      Say hello from me.

    8. Re:Imagine by Anonymous Coward · · Score: 0

      Imagine driving your electric shark through the city streets, a scream after midnight, stealing wireless power and pwning switching routers as your taillights disappear into the fog...

    9. Re:Imagine by ThatAblaze · · Score: 1

      tl;dr? In short the wacky network architecture has indeed been making vehicles harder to hack.

    10. Re:Imagine by Anonymous Coward · · Score: 0

      You mean Uber?

    11. Re:Imagine by Pentium100 · · Score: 2

      Do not want. All this complexity of modern cars makes me appreciate my classic car (no computers at all) more and more.

    12. Re:Imagine by Balp · · Score: 1

      The fibers are mostly, almost only used for the MOST bus, The most bus is either 25, 50 or 150 that is rought about he MBaud speed. the 25 and 150 can use fiber and the 50 and 150 uses electric contacts. I think most new cars with most uses electric connectors. At least that's the case with the newest Cadillacs. Using fibre have many problems they are expensive. For the next generations infotainment i think most, if not all automakers have been looking at using something simplar and more standard like ethernet for years. Because of the special environment i the car, it will probably be a modified ethernet. E.g. BroadR (http://www.opensig.org/). GM is part of the Open sig, prompters group, so they will probably also use this. And the starts will be in infotainment, as this need more bandwidth.

    13. Re:Imagine by Anonymous Coward · · Score: 0

      Consumer grade ethernet has had optocouplers for ages. The only time I saw a damage in the last 15 years was a 120m (yes it works) cable hung over a street with a powerful storm, and it only destroyed one port of a cheap netgear switch.

    14. Re:Imagine by VortexCortex · · Score: 1

      "For example, imagine a driver getting turn-by-turn navigation while a front-seat passenger streams music from the Internet, and each back-seat passenger watches streaming videos on separate displays."

      Imagine!

      Except they're already doing it now on their fondleslabs.

      Are you implying that a vehicle's electronics could instead will be wireless? That's Insane! Why, next you'll be telling me that they'll get rid of induction energized tracks or rails by letting cars carry their own energy. Oops, sorry, I must have iterated too far and wrapped around the historic recursion point, AKA Unsigned Technology Overflow.

    15. Re:Imagine by AmiMoJo · · Score: 1

      I have a dream. A dream where cars have standard APIs for things like climate control, steering wheel controls and entertainment systems. A dream where I can buy compatible equipment and expect everything to work as well as the manufacturer's own equipment.

      Not gonna happen, obviously.

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    16. Re:Imagine by mjwx · · Score: 1

      What a fantastic idea! By switching to a well understood standard it will finally make our vehicles trivial to hack!

      Ohh.. wait.. maybe that isn't such a good thing? *shrug*

      Erm...

      Ethernet can be easily made a closed system like CAN is.

      Also CAN is already easily hacked... However you need to get physical access to the CAN BUS which either involves a lot of fucking about in the engine bay or plugging into the OBD2 socket in the drivers footwell. Something tells me an Ethernet replacement for CAN would be the same, but swapping the OBD2 for RJ45.

      However for most cars, to actually change anything you need to replace the ECU or at the very least, the ECU firmware (car enthusiasts have been "hacking" their engines for decades now).

      --
      Calling someone a "hater" only means you can not rationally rebut their argument.
    17. Re:Imagine by ThatAblaze · · Score: 1

      Your lack of imagination does not make a thing impossible. The radio, in most cases, has direct access to the CAN bus. Do you think every radio on the market has vetted their firmware for buffer over run exploits? The implementation of privileges on a CAN bus is a joke, as you can see in the video I posted a couple of posts up. The biggest impediments to someone writing a script that can take over your car are 1) There is very little money in it and 2) None of the CAN busses follow a standard protocol and each car manufacturer seems to just throw together whatever implementation they feel like.

      Adapting an industry-wide standard would at least eliminate #2.

    18. Re:Imagine by mjwx · · Score: 1

      Your lack of imagination does not make a thing impossible.

      You're lack of understanding does not make your fantasy exploits real.

      None of the CAN busses follow a standard protocol

      And here you demonstrate you don't know what CAN bus is. It's a bus, not a protocol. CAN bus specifies the physical layer and encapsulation, but not the content and it's the content that differs between manufacturers.

      It's like using FTP and Bit Torrent, both run over ethernet but that doesn't make them compatible.

      Changing from CAN bus to ethernet wont change a thing except for speed and cost, because Ethernet is faster and cheaper. Oh and your OBD2 port might become a RJ45... thats about it. Vendors will continue to use their own protocols on a different network technology.

      The radio, in most cases, has direct access to the CAN bus.

      The radio often does not connect to the same CAN bus as the MAF sensor or ECU and the amount of information available to it is limited. Beyond that, the ECU itself needs to be modified before you can make any real changes.

      You've never modified car. No, I'm not asking, it's obvious that you've never even googled it if you think a buffer exploit from a head unit is a good way to do some damage (such as masquerading as a MAF sensor).

      The biggest impediments to someone writing a script that can take over your car are

      1) Implementation requires physical access.

      This is such a big one, it's almost a waste of time continuing.

      2) In most cases, you need to car to be modified before an exploit becomes dangerous (yep, you could bugger around with the radio, or maybe the seat adjustment, but not anything important like turbo engagement, throttle response or AWD systems).
      3) the sheer number of different types of systems make a large scale exploit extremely difficult.

      There's plenty of financial incentive from law enforcement, militaries and insurance companies.

      It sounds like you've been watching too many Fast and the Furious movies. You cant simply fire a grapple at a car and override all the electronics. Your medication, sounds like you need two of these and you need them by tonight.

      --
      Calling someone a "hater" only means you can not rationally rebut their argument.
    19. Re:Imagine by ThatAblaze · · Score: 1

      You've obviously been spending all your time working with manufacturer approved devices in the manufacturer's API. Escalating privilege is trivial if you don't rely on the manufacturer's API.

      So 1) Not true 2) Not true 3) Was my original point. Anything that standardizes the interface just serves to make things easier to hack.

      Your baseless attempts at character assassination just make you sound desperate.

  3. What?? by plebeian · · Score: 5, Insightful

    Lucas123 wants to stream audio and video across the same switches as his throttle by wire?????, I say we sell tickets to this event!

    --
    "I myself am made entirely of flaws, stitched together with good intentions."
    1. Re:What?? by amiga3D · · Score: 0, Redundant

      Really. What could possibly go wrong?

    2. Re:What?? by amorsen · · Score: 4, Insightful

      Mixing entertainment systems and critical safety systems on the same bus is common already. The only change is that with ethernet you get decent bandwidth and well-understood QoS.

      --
      Finally! A year of moderation! Ready for 2019?
    3. Re:What?? by kelemvor4 · · Score: 2

      Mixing entertainment systems and critical safety systems on the same bus is common already. The only change is that with ethernet you get decent bandwidth and well-understood QoS.

      QoS is OSI level 7. Ethernet is OSI level 1. There is no reason to assume that TCP/IP or QoS will be standardized upon or even used at all here.

      Also, QoS is a total dog if you are trying to employ it on consumer grade equipment. At least, that's been my experience with numerous linksys, d-link, and netgear devices. I'm kind of down on QoS as a result. Great idea, ruined by the implementation that most consumers will ever see.

    4. Re:What?? by SJHillman · · Score: 1

      You may want to review your OSI model. I don't remember QoS, but layer 1 is the physical medium. Ethernet is layer 2.

    5. Re:What?? by LordLimecat · · Score: 2

      Layer 7 is application layer, which is indeed where QoS sits (it deals with the source of the data).
      Ethernet, as in 802.3 is indeed layer 2, but wikipedia confirms: its more than just that. It includes the physical layer too.

    6. Re:What?? by Anonymous Coward · · Score: 0

      Mixing entertainment systems and critical safety systems on the same bus is common already. The only change is that with ethernet you get decent bandwidth and well-understood QoS.

      QoS is OSI level 7. Ethernet is OSI level 1. There is no reason to assume that TCP/IP or QoS will be standardized upon or even used at all here.

      Right. And the overwhelming majority of vendors itching to put their ethernet-enabled devices in cars without having to re-write an entire protocol stack WILL push for TCP/IP. I can see that shit on the wall now, as they all bitch in unison about the additional costs of re-tooling comms.

      And security will go right out the window, as it usually does when standing in the way of greed.

    7. Re:What?? by Anonymous+Psychopath · · Score: 1

      You may want to review your OSI model. I don't remember QoS, but layer 1 is the physical medium. Ethernet is layer 2.

      And Diffserv operates at layer 3, assuming TCP/IP. QoS definitely does not take place only or even primarily at layer 7.

      --

      Eagles may soar, but weasels don't get sucked into jet engines.

    8. Re:What?? by Anonymous Coward · · Score: 0

      I don't see anything wrong with this. Just build the whole system up to the standards of throttle/steer by wire and run the traffic on separate vlans.

      The next big thing is that is about to debut is OBD to be run off bluetooth and/or wireless (connected to tablets) rather than a proprietary cable / designated device. As soon as the car pulls into the service bay, it communicates with the tablet with full diagnostics and vehicle repair history. There is a fast-growing car company that is about to release a model with this capability in coming weeks. I'm not sure if TSLA does this already, but it isn't them. It was developed in Asia but it is being rolled out in the US.

      It may seem like no big deal, but it will be a huge leap forward in the automotive sector, which still operates the same as they did 20 years ago in many respects. The diagnostic data will be updated frequently, with diagnostics tied with technician repair strategies and outcomes. I.e. cylinder 4 running a bit hot? Well, a mechanic in Toledo and another in Miami found that when this was happening, this specific part was causing it. The control center reviews all of the repair histories and compiles the data, then suggestions are automatically sent to the mechanic when the problem is detected. This will lead to a faster repair, and looking at the bigger picture it will give more accurate data to the manufacturers of OEM replacement parts as to what needs to be produced and shipped (sooner). Faster repair under warranty = cheaper for the manufacturer and a happier customer.

      Unfortunately the service bays and dealerships aren't ready yet, with the exception of one. But once it gets going, it will save a great deal of time and money and may lead to better customer satisfaction.

    9. Re:What?? by Anonymous Coward · · Score: 0

      I'm not sure which OSI level this would correspond to, but on CAN, priority is all the way down at the physical level. There are active and passive bits on the bus. The active bit wins in a collision, and the higher priority packet goes out without having to be re-transmitted. The lower priority packet is then re-sent when the bus is open again.

      I don't want to be in a vehicle where the high priority traffic gets priority over entertainment 99.99% of the time, and my brake command gets delayed the other 0.01%...

    10. Re:What?? by Anonymous Coward · · Score: 0

      right. and chocolate is layer 4!

    11. Re:What?? by plebeian · · Score: 3, Interesting

      As a systems/network administrator I must say that If you are relying on general purpose wan connection for life or death services you are doing it wrong. Where I work we physically segment everything that is truly critical. The fire and alarm systems have multiple redundant connections including two that are 100% separate from our data network. The closest thing we have to a critical system running on a general purpose network is the use of SIP to provide connections from our phones to the PBX and that system has had a number of minor problems in the 7 years we have been using it. Ultimately if a phone call gets dropped in an office building the chances of someone dying because of it are truly minuscule. If on the other hand a drive by wire function fails you have a lot larger chance of death. I believe they will segment mission critical systems to a dedicate physical bus with redundant links in any proposed in car network. That way a entertainment system cannot interfere with the operation of say the headlights. My comment was made to expose the naivete of the original post and not to offer any truly insightful criticism.

      --
      "I myself am made entirely of flaws, stitched together with good intentions."
    12. Re:What?? by Anonymous Coward · · Score: 0

      Yeah right. Most modern vehicles and other large machinery already runs multiple busses in parallel (I should know, I work on one). It would be trivial, even the most logical design choice, to keep realtime functions on CAN and put 'courtesy functions' on ethernet..

    13. Re:What?? by Anonymous Coward · · Score: 0

      There is no technical reason that an automotive deployment could not use custom switching ASICs to multiplex traffic from different priority levels and have dedicated priority queue buffers to guarantee whatever delivery scheme you want.

      You could tolerate some well-defined maximum delay for the priority data frame to wait for an earlier frame to finish transmitting before the priority stream starts monopolizing the link. You would build your topology with a maximum path length to manage the maximum path delay for high priority traffic based on the number of links and switching buffers the frame might have to traverse. You would guarantee that the priority traffic would never cause itself congestion and allow lower priority traffic to use best-effort of idle capacity only. You can allocate enough hardware buffers to support all your important priority traffic getting buffered and scheduled consistently, since you can design the whole priority system as a closed world.

      The nice thing about gigabit would be the possibility to over-provision enough to allow a nice simple backbone topology through the car with redundant/failover links. Imagine just cat 5 and power cables running fore and aft, with all other switching and signalling being done by packet-based communication. You could even use something like power-over-ethernet for the many low-power peripheral controllers, though I imagine you'd need some other rugged connector format and you might be better off with DC power and fiber communications.

    14. Re:What?? by skids · · Score: 1

      It's called 802.1p, a mechanism for QoS tagging in a dot1q tagged frame.

      That said, this move could give new meaning to the second C in CSMA/CD.

    15. Re:What?? by fuzzyfuzzyfungus · · Score: 1

      Really. What could possibly go wrong?

      Luckily, ethernet totally has this covered: in the event of a collision, you just back off for a random number of milliseconds and then retransmit. No big deal! And cars are basically just big packets, right?

    16. Re:What?? by Jeremi · · Score: 1

      Also, QoS is a total dog if you are trying to employ it on consumer grade equipment. At least, that's been my experience with numerous linksys, d-link, and netgear devices. I'm kind of down on QoS as a result.

      Automobile manufacturers are considering using AVB as the mechanism by which they get their quality-of-service guarantees -- basically re-using the audio/video bandwidth-reservation protocols as a way to reserve bandwidth for their command signaling data.

      Whether or not this is a good idea I will leave as an exercise to the reader; but at least it is not relying on your father's broken QoS system.

      --


      I don't care if it's 90,000 hectares. That lake was not my doing.
    17. Re:What?? by epyT-R · · Score: 4, Insightful

      No thanks. I don't want to deal with my car getting hacked/stolen/monitored/remote controlled, which is infinitely more likely than this overwrought system.. I don't mind it for medical care, but not for my car. Cars should be stupid simple.

    18. Re: What?? by Anonymous Coward · · Score: 0

      You've been able to get Bluetooth obd-ii adapters for quite some time and can link to applications on a laptop or mobile device (like torque for android)

    19. Re:What?? by gewalker · · Score: 1

      I suppose, from a quantum mechanics viewpoint, cars cans be considered very large packets indeed. But quantum mechanics will still screw you over and fix your broken framitz if they think they can get away with it.

    20. Re:What?? by putaro · · Score: 1

      Just because it's done doesn't make it a good idea.

    21. Re: What?? by Anonymous Coward · · Score: 0

      I use custard as layer 4, you insensitive clod!

    22. Re:What?? by amorsen · · Score: 1

      There is no reason to assume that TCP/IP or QoS will be standardized upon or even used at all here.

      There is every reason to assume that. The car manufacturers are working hard not just to standardize on IPv6 in general, but in fact to have a common approach to such things as address allocation. QoS will be much easier to handle with ethernet, not because it is less complex but because the code is already written and widely deployed.

      Also, QoS is a total dog if you are trying to employ it on consumer grade equipment.

      I must admit that I have never tried to use QoS on ethernet with consumer grade equipment. Why would you want to though? Generally you have precisely one switch at home, and that switch is typically capable of simultaneous full speed on all ports, so it only drops packets if multiple input ports are trying to send more than 1 gigabit in total to one output port. I have difficult imagining that scenario in a home.

      QoS on the WAN is entirely different, but the WAN is typically not yet ethernet, or at least not ethernet at standard speeds.

      --
      Finally! A year of moderation! Ready for 2019?
    23. Re:What?? by amorsen · · Score: 1

      I believe they will segment mission critical systems to a dedicate physical bus with redundant links in any proposed in car network.

      I will be surprised if they do that. It would make sense, but since they do not do that today, why should they suddenly start doing so?

      --
      Finally! A year of moderation! Ready for 2019?
    24. Re:What?? by Anonymous Coward · · Score: 0

      QoS is OSI level 7.

      But who uses OSI anyway? I for one have never seen a system using an actual implementation of the OSI stack.
       
       

      Ethernet is OSI level 1.

      No it is not. OSI level 1 is a totally different protocol, which to the best of my knowledge is used by no-one.

    25. Re:What?? by Bing+Tsher+E · · Score: 1

      They will do that because if it comes up as a safety issue in a casual discussion on Slashdot, it will occur to the engineers designing the system. Let's not get carried away that these threads on Slashdot are sessions at an actual for-real Genius Bar.

    26. Re:What?? by msauve · · Score: 1

      No, QoS is not L7. There's QoS available at L2 (802.1p/q), and also at L3 (diffserve/intserv).

      --
      "National Security is the chief cause of national insecurity." - Celine's First Law
    27. Re:What?? by kelemvor4 · · Score: 1

      There is no reason to assume that TCP/IP or QoS will be standardized upon or even used at all here.

      There is every reason to assume that. The car manufacturers are working hard not just to standardize on IPv6 in general, but in fact to have a common approach to such things as address allocation. QoS will be much easier to handle with ethernet, not because it is less complex but because the code is already written and widely deployed.

      Also, QoS is a total dog if you are trying to employ it on consumer grade equipment.

      I must admit that I have never tried to use QoS on ethernet with consumer grade equipment. Why would you want to though? Generally you have precisely one switch at home, and that switch is typically capable of simultaneous full speed on all ports, so it only drops packets if multiple input ports are trying to send more than 1 gigabit in total to one output port. I have difficult imagining that scenario in a home.

      QoS on the WAN is entirely different, but the WAN is typically not yet ethernet, or at least not ethernet at standard speeds.

      Perhaps FIOS is still atypical. 300Mbps/75Mbps is what I've got. Pity I can't get 300/300.

    28. Re:What?? by plover · · Score: 1

      Cars should be stupid simple.

      Except cars work at high speeds in varying and complex environments, and are often driven by stupid, simple, frail, inattentive, incapacitated, and/or incompetent people. A smart system can help keep them in control of their vehicles, perhaps keeping them from spinning out in front of you, or driving through an intersection into you. A smart system can understand how fast you're moving, how much inertia you have, and whether it should deploy the airbags to save you. It can detect wheel slippage, and use the ABS system to keep you moving straight. A smart system can monitor combustion and modify engine performance to reduce emissions, or shift the transmission to improve fuel economy. A smart system can call emergency services in the case of an accident. A smart car can adjust the suspension to adapt to rough or smooth roads.

      --
      John
    29. Re:What?? by amorsen · · Score: 1

      Perhaps FIOS is still atypical. 300Mbps/75Mbps is what I've got. Pity I can't get 300/300.

      That is not a counterexample. 300Mpbs or 75Mbps are not standard ethernet speeds. To do QoS at those speeds you are either at the mercy of the provider-controlled CPE or you try to shape your traffic in your own router. Either way you are into territory where you can make it work, at least with some routers.

      This is very different from the situation at 100Mbps or 1Gbps where QoS tends to work just fine without special configuration, except for enabling it on those devices that have it disabled in the default configuration.

      --
      Finally! A year of moderation! Ready for 2019?
  4. The Ether! by Anonymous Coward · · Score: 0

    Ethernet is found everywhere – on mobile phones, desktop computers, Blu-ray Disc players, set top boxes, and even in your car.

    1. Re:The Ether! by cgimusic · · Score: 1

      Will we be asked to install the Ask Toolbar every time we plug in an Ethernet cable too?

  5. Stop reinventing the wheel. by Anonymous Coward · · Score: 0

    Why did these morons spend time and energy to create LIN, CAN/CAN-FD, MOST and FlexRay in the first place?

    1. Re:Stop reinventing the wheel. by SJHillman · · Score: 1

      For the same reason they invented IPX/SPX, Appletalk and TCP/IP in the first place? Or for the same reason they invented Beta and VHS? Or Blu-Ray and whatever that other one was? There have always been competing standards.

    2. Re: Stop reinventing the wheel. by jxander · · Score: 4, Funny
      --
      This signature is false.
    3. Re:Stop reinventing the wheel. by Sanhedran · · Score: 2

      Why did these morons spend time and energy to create ... CAN/CAN-FD

      If you think the CAN standards were developed by "morons", there's no educating you.

  6. Can the engineers please tell programmers to GTFO by Anonymous Coward · · Score: 1

    This is serious stuff. You don't want the brakes to stop working just because someone decides to start a Bittorrent client.

  7. Micro USB by locopuyo · · Score: 1

    They should just make it Micro USB. Then you can charge your electric car and transfer data with the same cable. Just like a phone. I don't want to have to buy new cables just so I can charge my car.

    1. Re:Micro USB by Anonymous Coward · · Score: 0

      Cables? What about chargers? Why can't I charge my electric car with my phone's charger? This needs to change!

    2. Re:Micro USB by Anonymous+Psychopath · · Score: 2

      Cables? What about chargers? Why can't I charge my electric car with my phone's charger? This needs to change!

      Can confirm unable to charge PEV with my laptop's USB port. Epic fail on the part of the manufacturer.

      --

      Eagles may soar, but weasels don't get sucked into jet engines.

    3. Re:Micro USB by Arker · · Score: 1

      You know, that's a great idea, except for how incredibly long it would take to charge your car!

      --
      =-=-=-=-=-=-=-=-=-=-=-=-=-=-
      Friends don't let friends enable ecmascript.
    4. Re:Micro USB by gstoddart · · Score: 1

      Bah, just use one of the 2.1A USB things instead of the 1A USB things and it'll be fine, right? ;-)

      --
      Lost at C:>. Found at C.
    5. Re:Micro USB by Immerman · · Score: 1

      Seriously man. I mean my laptop can stream like GB/s, why can't my car that only draws a few KW charge from it? Like seriously, dude. It should be able to run for, like, thousands of hours. Kilo versus Giga baybe.

      --
      --- Most topics have many sides worth arguing, allow me to take one opposite you.
    6. Re:Micro USB by Bing+Tsher+E · · Score: 1

      The trouble is, you're thinking in Metric.

      Think Gallons of fuel. It'll keep you out of trouble.

    7. Re:Micro USB by Neil+Boekend · · Score: 1

      Next step is 2.1 MA over superconducting micro USB cables to fix that little issue.

      --
      Well, I might have a way, but it only works on a semi spherical planet in a vacuum.
  8. A little late by zerofoo · · Score: 1

    The SAE has been talking about this for years. This article is from 2012:

    http://articles.sae.org/11142/

    I think the BMW 7 series has used ethernet for the infotainment systems for a while now.

    1. Re:A little late by LordLimecat · · Score: 1

      Ethernet seems like the wrong system to be using for anything where the configuration is fairly static and reliability is at a premium. You really want your steering wheel traffic hitting the windshield wipers because their NIC blew up and started claiming every MAC address as their own?

      You can fix this stuff on enterprise networks with really expensive switches with L2 authentication / sticky MACs, STP, etc-- but that seems like the definition of "overengineered".

    2. Re:A little late by drinkypoo · · Score: 1

      Ethernet seems like the wrong system to be using for anything where the configuration is fairly static and reliability is at a premium. You really want your steering wheel traffic hitting the windshield wipers because their NIC blew up and started claiming every MAC address as their own?

      What makes you think this can't happen with CAN bus?

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    3. Re:A little late by epyT-R · · Score: 1

      It wont' happen if you use a couple of switches and some relays for the wipers instead, and mechanics for the wheel/accel/brake etc....a lot cheaper too.

    4. Re:A little late by Immerman · · Score: 1

      Ethernet = several decades of bug-fixes against everything malicious SOBs can throw at it, included in all but the cheapest off-the-shelf networking hardware. CAN, etc = ???.

      Reinventing the wheel is usually a bad idea unless you're shooting for something fundamentally different than anything that has been attempted before.

      --
      --- Most topics have many sides worth arguing, allow me to take one opposite you.
    5. Re:A little late by fluffy99 · · Score: 1

      It wont' happen if you use a couple of switches and some relays for the wipers instead, and mechanics for the wheel/accel/brake etc....a lot cheaper too.

      But then you can't have a smart car with a moisture sensor and rain detector to automagically turn the wipers on for you. Although, I have gotten spoiled by not having to remember to turn on/off the headlights. Same deal for interior lights, - you could go with the old school mechanical switches but it is nice to have them turn on at the appropriate times and turn them selves off if your toddler left the light on and you didn't notice.

      Brakes and steering are still mechanical, btw.

    6. Re:A little late by Anonymous Coward · · Score: 0

      Mechanical relays that wear out and are a lot more expensive because they involve more assembly, or solid state relays that can die randomly due to voltage spikes even with massive over-voltage and surge protection? And at some point, although maybe not so much for cars, a bus is cheaper than simple individual wires because of the amount of copper and assembly labor saved.

    7. Re:A little late by Balp · · Score: 1

      We are talking Ethernet BroadR here, the cabling are different. Any car design if probably still going to need multiple networks.

    8. Re:A little late by Balp · · Score: 1

      You have any idea how much cable would be need ed in a car of today with that kind of wirering. How that would make the car a bit heavier. and how that would make the car work? I heard numbers form Volvo, moving form switches and relays to can busses made them remove about 200 kilos in the car. While adding functionality.

    9. Re:A little late by mirix · · Score: 1

      CAN is just as old as ethernet, if that is your test of reliability.

      --
      Sent from my PDP-11
    10. Re:A little late by LordLimecat · · Score: 1

      Im not familiar with CAN bus, but ethernet has a lot of behaviors that are necessitated by its dynamic nature. For instance, devices ask "who has ethernet address ABC" as a broadcast, and whoever responds "I DO" is recorded by the endpoint and the switch for routing.

      That works great when your configuration changes a lot and you need someone to be able to plug a laptop into the network and communicate with the router without knowing its MAC ahead of time. It works really poorly when someone misbehaving on the network can cause your car to malfunction and / or crash.

    11. Re:A little late by LordLimecat · · Score: 1

      "All but the cheapest hardware" means youre plunking down $500+ for a midrange switch in your car that runs its own operating system, handles VLANing, and does sticky MAC-- all to run the internals in your car.

      The more complicated you make something, the more likely it is to break. That idea would make Rube Goldberg proud.

    12. Re:A little late by Neil+Boekend · · Score: 1

      Not having to turn on the lights without electronics is easy. If the car is running the lights should be on anyway. Darkness has nothing to do with it, as even during the day a car with the lights on is far more easily seen.

      --
      Well, I might have a way, but it only works on a semi spherical planet in a vacuum.
  9. Re:Don't quit your day job by roc97007 · · Score: 2

    Yeah, everyone knows the Cylons are going to show up any day now.

    --
    Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
  10. car bandwidth needs by Anonymous Coward · · Score: 0

    What exactly do you need 100 or 1000 Mbit/s in a car for? (Especially in the control system.)

    Are we talking about having the engine control on the same bus as the entertainment system (where playing a video/DVD could use a lot of bits)?

    1. Re:car bandwidth needs by Balp · · Score: 2

      First all that have video transmission, as parking assistance, nightvision, review cameras. Infotainment. I don't think the same bus, but in the future using the same technology. One technology, one open standard might bake stuff overall easier to implement, and that lowers costs or increases quality, or features, or a mix of these.

  11. Cylons by Anonymous Coward · · Score: 0

    That will be the least of our concerns.

  12. And it'll be... by roc97007 · · Score: 1

    IPV4. bets?

    --
    Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
    1. Re:And it'll be... by decsnake · · Score: 1

      nah, IPX

    2. Re:And it'll be... by roc97007 · · Score: 1

      nah, IPX

      (slaps forehead) Of course!

      --
      Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
    3. Re:And it'll be... by LordLimecat · · Score: 1

      Irrelevant on a private network.

    4. Re:And it'll be... by grim4593 · · Score: 1

      Not a chance! With the advent of car to car networking and internet accessibility every component in your car will have its own internet facing IPV6 address!
      Imagine the possibilities: You could open your trunk from anywhere in the world, you could preheat the car before your flight even lands, you could update your license plate in real time!

    5. Re:And it'll be... by roc97007 · · Score: 1

      Every component of my car outward facing.... that's absolutely terrifying.

      --
      Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
    6. Re:And it'll be... by swb · · Score: 1

      Even though we're all supposed to believe IPX was inferior to IP, I still think the addressing scheme has some advantages over IPv4.

      It's a larger address space, with 32 bits of network addressing and 48 bits of node addressing using the MAC address by default, and there was no need for DHCP as clients could easily autoconfig node addresses by listening to the wire.

      Assuming IP had adopted this addressing scheme, I sometimes wonder how many man-hours and dollars would have been saved over the years from the effort expended when merging or internetworking private IP networks with overlapping address spaces (renumbering or implementing NAT) or expanding networks that outstripped /24s.

      Would even be talking about IPv6 if the network space of IPv4 included 32 bits of network?

      A /16 is a huge allocation by existing IPv4 standards, but a relatively small total network in terms of /24s, yet with IPX style addressing a single /16 network assignment could satisfy the internetworking needs of a huge network since every bit could be used for a unique subnet, and fewer total subnets would be needed since the unique node space is the entire MAC space. LANs could scale without routing layers (up to whatever is sane for a single broadcast domain).

      For example, McDonalds has 34,000+ restaurants worldwide. A single /16 IPX-style network assignment could be used to internetwork every restaurant in the world with more than enough left over to cover every other possible use. A single /8 could cover squad-level unique network assignments to the entire People's Liberation Army and probably unique assignments to every single vehicle, ship and plane in inventory.

    7. Re:And it'll be... by roc97007 · · Score: 1

      Irrelevant on a private network.

      Well, yes. I wrote a comment on the article "why haven't we run out of IPV4 addresses yet" (the most recent article a few days ago, not the one last week or the one last January or the one around Thanksgiving, or the one late summer...) saying that the reason we haven't run out is that most of us are using NAT now, and have only one, or at least very few, actual outwards facing addresses. I strongly suspect that car components will have a similar arrangement. My comment above was a mild zinger to the hand-wringers who think every single device should be IPV6 RIGHT NOW.

      Parenthetically, the idea of every one of my devices having an unique, outward facing network address is absolutely terrifying.

      --
      Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
  13. People seem to be misunderstanding by kelemvor4 · · Score: 4, Informative

    This article is about the L1 PHY layer, not the L2 Data Link layer. There is no reason to assume this means your car will be using TCP/IP. The diagram in TFA clearly indicates that the PHY layer being discussed here is independent of the protocol.

    In fact, the included diagram seems to indicate broadcom is pitching some kind of adapter device which would enable inclusion of the new L1 layer with no changes whatsoever to the programming of the devices on either end. One would hope that such a thing would be only considered a stop-gap measure while they reworked their components to use the new bus natively in future models. History clearly shows that such adapters tend to be inefficient.

    1. Re:People seem to be misunderstanding by drinkypoo · · Score: 1

      This article is about the L1 PHY layer, not the L2 Data Link layer. There is no reason to assume this means your car will be using TCP/IP. The diagram in TFA clearly indicates that the PHY layer being discussed here is independent of the protocol.

      There's every reason to suspect that TCP/IP will be used for audio/video modules. There's no particular benefit to using anything else; no matter what, crypto is going to have to be a part of it.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    2. Re:People seem to be misunderstanding by jrumney · · Score: 2

      There's every reason to suspect that TCP/IP will be used for audio/video modules. There's no particular benefit to using anything else

      Except that using protocols that are designed for audio and video instead of general IP traffic ensures that you see/hear a glitch free stream without waiting for buffering before you start playback every time.

    3. Re:People seem to be misunderstanding by TheSync · · Score: 1

      The article mentions Parrot's Ethernet AVB connected systems. The carriage of audio/video media over AVB has been standardized by the AVB Transport Protocol in IEEE 1722, and yes, it is just Ethernet, no IP.

      The theory is that your car is a LAN, and does not need to have Layer 3.

    4. Re:People seem to be misunderstanding by Blaskowicz · · Score: 1

      Funny, with digital TV, any kind of it including over-the-air, you have to wait a few seconds for buffering. (or waiting for the next full frame and whatever, I don't know in great detail and accuracy)
      It was quite noticeable when it coexisted with analog TV and channel switching was instantaneous on that one.

      Also the parent (or grand-parent) forgot about UDP/IP, to niptick a bit.

    5. Re:People seem to be misunderstanding by jrumney · · Score: 1

      MPEG video streams are slow to start because not every frame is a full frame. There are also latencies everywhere within the transmission system of digital broadcasts that adds up to a couple of seconds delay when compared side by side with analog broadcasts. But the additional buffering required to get smooth playback on a shared TCP connection would be multiples of that delay.

  14. No by Tailhook · · Score: 4, Interesting

    If it's good enough for commercial aircraft it's good enough for your car.

    --
    Maw! Fire up the karma burner!
    1. Re:No by Anonymous Coward · · Score: 2

      No, it really isn't. Commercial aircraft aren't traveling a few seconds' distance behind other aircraft.

    2. Re:No by 140Mandak262Jamuna · · Score: 2, Insightful

      No it is not good enough. Commercial aircraft is flown by people with constant training, check lists, protocols, under the guidance of air traffic control. They are supposed to be not drunk, supposed to be well rested. Cars? driven by everyone from pimply teens giggling and texting while driving all the way up to 90 year old grandma who only has a vague nebulous feedback from her right foot when she is on arthritis medication.

      --
      sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
    3. Re:No by Immerman · · Score: 4, Interesting

      And? The network between the brake pedal and the brake doesn't give a flying fsck about the state of the driver, just make abso-%$#@!-ing-lutely sure that nothing the user (or a malfunctioning/malicious app ) can interfere with the signal. For starters don't put anything user-accessible on the same network - insert a heavily firewalled router at least, and preferably an old-fashioned air-gap.

      --
      --- Most topics have many sides worth arguing, allow me to take one opposite you.
    4. Re:No by fluffy99 · · Score: 1

      The only thing AFDX has in common with ethernet is the mac layer. It's incompatible with and looks nothing like standard tcp/udp you normally see running around on ethernet nowadays.

    5. Re:No by Balp · · Score: 1

      TCP/IP standard have nothing to do with ethernet, even if they match well. We are using TCP/IP on many other physical layers as well.

      I think most automakers think about adding tcp/ip at the same time as ethernet thou. Just becase it have extra values.

  15. ethernet =/= internet by Culture20 · · Score: 1

    imagine a driver getting turn-by-turn navigation while a front-seat passenger streams music from the Internet, and each back-seat passenger watches streaming videos on separate displays.

    Just because there's an internal network for the car's electronics doesn't mean there's any internet connection (and there'd better not be).

    1. Re:ethernet =/= internet by Anonymous+Psychopath · · Score: 1

      imagine a driver getting turn-by-turn navigation while a front-seat passenger streams music from the Internet, and each back-seat passenger watches streaming videos on separate displays.

      Just because there's an internal network for the car's electronics doesn't mean there's any internet connection (and there'd better not be).

      Just the possibility of playing NetWars on my car's intranet has me all in a tizzy.

      --

      Eagles may soar, but weasels don't get sucked into jet engines.

    2. Re:ethernet =/= internet by bob_super · · Score: 1

      But there will be. Because some marketing genius will convince the higher-ups that it will save $50 per car on the implementation of On-Star.

      I am still waiting for the great-On-Star-Hack-megaJam. We should make it easier for someone to shutdown half the US cars in one mouse-click...

  16. and be locked into the poor build in radio system by Joe_Dragon · · Score: 3, Insightful

    and be locked into the poor build in radio system that can't be upgraded to a better 3rd part one.

  17. Not mine by Anonymous Coward · · Score: 0

    I will never buy a car that doesn't have a hard maximum in the time it takes between when I hit the brakes and the actual brakes engage. Ethernet is good for a lot of things, but there are many systems where its drawbacks are just too big, and every part of controlling a car falls in the latter category. Otoh, if they want to implement it all using token ring, maybe we can talk.

    1. Re:Not mine by petermgreen · · Score: 1

      Otoh, if they want to implement it all using token ring, maybe we can talk.

      Modern ethernet networks are a network of switched full duplex point to point links. CSMA/CD is still supported in the general purpose copper physical layers but only for backwards compatibility reasons.

      --
      note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
  18. AVC-LAN by adiposity · · Score: 1

    What about AVC-LAN (what Toyota uses)?

    Not that I would suggest it has the bandwidth of ethernet!

    1. Re:AVC-LAN by jrumney · · Score: 1

      AVC-LAN is even slower than CAN. Its basically a remote control bus for the A/V components to send commands to each other over. Actual audio and video is taking a different route (probably analog).

  19. Leashes? by Anonymous Coward · · Score: 0

    Is anyone else wondering what the hell the poster actually means? Are the cars going to run blue cables back to houses?

  20. WTB IPv6 by dave562 · · Score: 2

    Need teh NATz for my car!

  21. Your Next Car's Electronics Will Likely Be Connect by Anonymous Coward · · Score: 0

    Was I the only one who had a moment where they were trying to envision the car being connected to a really really really long cat 6 cable to your house?

  22. Back seat passengers ... by PPH · · Score: 1

    ... watch their streaming video on their phones and tablets. That way, they can carry their downloaded content with them away from the car.

    Built-in entertainment systems are stuck in the vehicle. And unless you have a service that allows it, they require additional service plans for those devices. A phone/tablet needs (for longer trips) a USB power port. A 12 volt port with suitable adapter does just fine.

    --
    Have gnu, will travel.
    1. Re:Back seat passengers ... by Blaskowicz · · Score: 1

      The way I'd do it would be a file server with 1TB+ hard drive, that'd be better than a puny 16GB flash per tablet (assuming that tablet can access file shares on a network, which probably leaves Apple stuff out)
      USB power jacks built-in directly to the car with the latest "Power Delivery" specifications would be nice as well, no need for 12V outlets (round? cigar plug?).

      Or you know, passengers could sleep, watch the scenery, talk to each other.

  23. Hmm by Ol+Olsoc · · Score: 2

    Will they provide the paperclip to reset my car when it hangs up?

    --
    The shepherds did so well protecting the flock that the sheep no longer believed that wolves existed.
  24. Re:and be locked into the poor build in radio syst by jrumney · · Score: 1

    If its standardized instead of the mess of proprietary buses we have now, it will reopen the possibility of replacing components.

  25. Duplex by Anonymous Coward · · Score: 0

    If they use Ethernet in my car that is fine, but it better damn well be full duplex as most drivers would prefer to avoid collisions. Almost worse than your wife jabbering away in the passenger seat with a set of runts in the back seat.

  26. ADHD by Lawrence_Bird · · Score: 1

    What is wrong with you people? Can't you just read a book or magazine or sleep if you aren't driving? Look out the window? Count the blue cars? An hour in solitary and you all will be curled in a ball moaning for mommy.

    1. Re:ADHD by Firethorn · · Score: 1

      Can't you just read a book or magazine or sleep if you aren't driving?

      Sometimes. Sometimes it makes me nauseous. Looking out the window only works for so long.

      --
      I don't read AC A human right
  27. I was promised a flying car by alzaid.saud · · Score: 1

    and the best they could do was ethernet ..... goddammit I wish the autoindustry would at least try to keep up with the times.

  28. Not for me, period. by Anonymous Coward · · Score: 0

    All new cars are hideously complex with respect to the electrical
    and electronic systems, and this stuff WILL lead to reliability problems.

    If you can buy a new car every two or three years, perhaps you will
    never have to deal with the problems, but the fact remains that the
    expense of diagnosis, repair, and replacement of some of the electronic
    systems in newer cars will exceed the value of the car after a relatively small
    number of years. This is a dream come true for the manufacturers, but a nightmare
    for those of us who don't want to be on the gerbil wheel of wasteful consumption when older
    cars which are well maintained can do the required job nicely.

    You can disagree with me, but I have a master's in mechanical engineering and a EE as well, and I have
    been repairing my own cars for over 40 years, so there is a pretty good chance I know
    what I am talking about.

  29. Hope they harden it. by Revek · · Score: 2

    Ethernet is notoriously susceptible to the emp from a close lighting strike. If you don't think so, just work in the cable industry for a while. After every serious lighting storm we will have several modems that appear fine except the ethernet is blown. It is usually the only thing burned out in the house. Often the rf side is still working fine and sending information back to the management system. It will suck when you have to go into a parts store and say gimme a box of ethernet chips for my car.

    1. Re:Hope they harden it. by Neil+Boekend · · Score: 1

      For important stuff the EMC demands are insane in cars. Usually parts are well protected from interference. Anything except a direct lightning strike isn't going to hurt it.
      If they switch to ethernet those demands are not going away. The infotainment systems may be susceptible, but the critical systems will not be.
      Besides that: why change everything to ethernet? Just install an ethernet system for non critical systems and keep the critical systems on CAN. They need to be as separate as possible anyway so why do they need to be the same type of network?

      --
      Well, I might have a way, but it only works on a semi spherical planet in a vacuum.
  30. Bandwidth by Anonymous Coward · · Score: 0

    Will you kids stop using up all of the brake and streering bandwidth?

  31. Yeah, but ... by gstoddart · · Score: 2

    For example, imagine a driver getting turn-by-turn navigation while a front-seat passenger streams music from the Internet, and each back-seat passenger watches streaming videos on separate displays.

    Now imagine how much the data plan for your car is going to cost you. You'll be locked into a plan with the car company and pay through the nose.

    No thanks. I have a dedicated GPS, an MP3 player I can connect to my car stereo, and most everybody has portable devices which can play video already.

    Now get off my damned lawn, because I don't want or need a car which is connected to the interwebs.

    --
    Lost at C:>. Found at C.
    1. Re:Yeah, but ... by Blaskowicz · · Score: 1

      Car companies are carriers now?
      I find this stuff unlikely.
      More likely, you change the car's nav and entertainment system's SIM card if you want to change cellular provider and if you don't want one, don't have any.

    2. Re:Yeah, but ... by gstoddart · · Score: 1

      More likely, you change the car's nav and entertainment system's SIM card if you want to change cellular provider and if you don't want one, don't have any.

      I have seen little evidence of car companies being willing to do such things in the things they make.

      Like every other company, they want lock in, monetization, and a big piece of the action of ongoing revenue.

      I find it hard to believe the won't try to force you into buying from them. They're not going to make things they aren't going to profit from, because there's nothing in it for them.

      They may not directly be carriers, but I'm quite certain they're likely to try to extract their cut and require a component they do control.

      --
      Lost at C:>. Found at C.
  32. Your next car will be a bike. by Hairy1 · · Score: 1

    News flash; we are coming to the end of the petrochemical age. We are very much at peak oil, and the way down will only see rising fuel costs. Buying a Hybrid may be more economic and efficient, but ultimately our whole way of life will be challenged. Get used to the idea that soon we will not have the pervasive availability of cheap fuel. Get on your bike.

  33. though... by DrYak · · Score: 1

    Nobody's saying "Man, I wish my CAN bus had more bandwidth so I could stream!

    Yup, in *theory* you know that a CAN bus is used for critical automotive functionality (say engine, ABS, power steering, or even drive-by-wire, autonomous steering, etc.)
    Whereas the streaming should stay confined within the media subsystem, and both should be kept completely isolated from each other.
    So it doesn't make sense to speak about successor of CAN bus technologies and media consumption in the infoteinment system of the car.
    They are completely separate networks.
    In theory.

    In practice, you know pretty much that we leave in a world of product rushed into production due to marketing constrain. A world where, due to extremely flacky design, it's possible to hack a vehicle by abusing the wireless transmission used to report tire pressure.
    So you know that lack of proper separation is bound to happens and you will end-up being able to hack a vehice by streaming a specially crafted video file, simply because the various ethernet networks aren't properly isolated from each other.

    --
    "Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
    1. Re:though... by dotbot · · Score: 1

      It's inevitable that the vehicle intranet will have a connection to the internet. Time to regulate the use of electronic systems for critical vehicle functions.

    2. Re:though... by Anonymous Coward · · Score: 0

      > it's possible to hack a vehicle by abusing the wireless transmission used to report tire pressure.

      No, it's not. It's possible to listen to the messages from the transmitters, clone the sensor ID and spoof false messages by basically overpowering the weak-by-design transmitters, and perhaps in some systems even get the light on the dash to come on. And unless the driver's an idiot, they'll at most pull to the side of the road to make a visual check. More likely, particularly if they're in a dodgy neighborhood, they'll drive on to see if the problem goes away on its own or decide that the tire is less important than getting to a more secure location before stopping. Hell, most people ignore the damned light for a week when it really is telling them a tire's low.

      As for "hacking a vehicle" -- no way. The TPMS receiver can encode messages which are likely transmitted on the CAN/J1850/whatever bus, but there's no way you're going to pass a corrupted message upstream from a tire sensor and somehow gain access to other systems.

      Safety critical systems in automotive applications are subjected to enormous levels of scrutiny, both by the auto makers and the tier one suppliers who provide them. They have to be, as the liability implications are huge. Infotainment systems may be at the whim of the marketers (though you have to keep in mind that every component in a car goes out the door with at least a three year warranty while being subjected to temperatures which would kill your consumer device), which is why the in-dash navigation system is so much more expensive than a Garmin that you stick to the windshield), but safety critical systems are fully under the auspices of people who truly know what they're doing.

      I'm no fan at all of touch screens in automotive control environments, and I fervently hope that some reason comes to the industry before these things replace sensible controls. It would be unfortunate if it takes the accidental death of some celebrity or high profile politician's innocent grandchild to restore sanity when it comes to essential controls, but given the path the industry is on, it may inevitably come to that. Let's not invalidate good technical arguments by blathering uninformed nonsense when you don't know what you're talking about.

      And if you're wondering, yes, I am a senior engineer in the automotive sensors business.

  34. Shared networking with user services? by Immerman · · Score: 4, Interesting

    What a horrible, horrible idea. Not the ethernet aspect, that makes sense, reinventing the wheel is usually a bad idea, and especially so when the competition has a multi-decade lead on eliminating bugs and malicious exploits and offers cheap, reliable off-the-shelf hardware. No, it's the idea of putting anything whatsoever user-accessible on the internal network I object to. If this data bus is carrying the information that tells my increasingly fly-by-wire care to apply the brakes or turn right to avoid oncomming semis then all it takes is one misbehaving flappy-bird clone spamming the network at the wrong moment to kill me, to say nothing of malicious attacks. There's absolutely no reason *anything* but internal systems communication should be on that network. Period. If you want an media network fine, but that can probably be provided far more cheaply and conveniently by including an airgapped $10 wireless hub with a 10' range that can only talk directly to things like the steering-wheel mounted media controls and the dashboard LCD/windshield HUD. And maybe a cellular modem. You're in a pretty decent approximation of a Faraday cage, so non-malicious outside interference should be minimal, and any communication with the mission-critical network should be heavily firewalled, at an absolute minimum. Not much reason to allow bi-directional communication at all - "spam" the wireless network with multicast up-to-the-second system and diagnostc data and you're good, at 0.01% of total bandwidth. No reason for anything not physically connected to be able to say a %$#@!* thing to the mission-critical components. If ever there was a non-hyperbolic use of the term "mission critical", maintaining control of a car is it.

    * %$#@! - when no variation of "fuck" is strong enough. Bonus points if you can pronounce it. Q-Bert did, but then he had that hose-nose to work with.

    --
    --- Most topics have many sides worth arguing, allow me to take one opposite you.
    1. Re:Shared networking with user services? by fluffy99 · · Score: 2

      There are multiple busses in vehicles already, separated by function. Engine controls are usually on a higher speed can bus, stuff like the speedo and body (lights, doors, etc) on a low speed can bus. I can see adding a third bus for entertainment type stuff such as the radio sat nav, wireless hotspot etc.

    2. Re:Shared networking with user services? by Anonymous Coward · · Score: 0

      reinventing the wheel is usually a bad idea

      I'd expect the car industry does employ some of the people who'd be best qualified for that task though.

    3. Re:Shared networking with user services? by VortexCortex · · Score: 2

      And then one token ring to rule them all, and in the blind spot bind them!

    4. Re:Shared networking with user services? by kyrsjo · · Score: 1

      I never hear of any vehicle where the breaks are actually fly-by-wire, neither the steering. Neither any manual transmission wich is fly-by-wire. Throtle and engine control yes, but I don't know if it's usually on the same bus as everything else, or on a separate network.

    5. Re:Shared networking with user services? by Immerman · · Score: 1

      Most every electric or hybrid vehicle uses brake-by-wire, you have to in order to get regenerative braking. They have normal friction brakes as well for emergencies, and given the number of power-related issues I've had over the years I would hope those are actually mechanically linked and just don't make contact until the pedal is almost floored, but I don't know that I would bet on it.

      Steering, I can't think of any examples off the top of my head, but have you not seen the "active wheel" projects being promoted by various manufacturers? Electric in-wheel drive, braking, steering, and suspension - all the essential mechanical elements of a car, ready to bolt onto a frame and be plugged in to the power and control systems.

      And while I couldn't swear to it, I believe I recall hearing that there is a trend towards unifying the various data buses, hence the need for an increasingly fast and responsive network.

      --
      --- Most topics have many sides worth arguing, allow me to take one opposite you.
    6. Re:Shared networking with user services? by wiredlogic · · Score: 1

      Critical engine management and safety functions are already controlled on isolated busses today. The people "hacking" their vehicles through the OBDII diagnostic port are not able to gain full access to all internal communication. Airbag systems for instance are controlled by their own dedicated low latency protocol on an isolated bus. This is necessary to guarantee that malfunctioning external equipment doesn't saturate safety critical busses with traffic. The proposals for high speed internal networks are limited to expendable features like entertainment, Wifi, navigation, and external cameras. If the network goes down you car will still keep driving.

      --
      I am becoming gerund, destroyer of verbs.
    7. Re:Shared networking with user services? by Anonymous Coward · · Score: 0

      You had me at Q-Bert. Bravo!

  35. I'm OK with ethernet in cars by PopeRatzo · · Score: 3, Funny

    As long as I don't have to make my own cables.

    One of the longest days of my life was many years ago when I told a friend I could wire up his little storefront business if he bought a spool of Cat5 and a bunch of connectors.

    I sat there with that crimping tool and my fumblefingers and invented entire new categories of curse words. A friend from a local Army base came by and for a few slices of pizza and a six-pack he knocked out those cables like nothing.

    It was a humbling experience. Which I probably should not have shared here on Slashdot because you guys were probably all making your own ethernet cables since your were like five years old.

    --
    You are welcome on my lawn.
    1. Re:I'm OK with ethernet in cars by dbIII · · Score: 1

      you guys were probably all making your own ethernet cables since your were like five years old.

      We're not all teenage fans of Justine Beaver here. It's not so long since co-ax was used for ethernet.

    2. Re:I'm OK with ethernet in cars by fwc · · Score: 1
      I was recently in Weird Stuff Warehouse in Sunnyvale. I like to walk through the store when I'm in the area just for the walk down memory lane... They have used computer stuff for sale from, well, pretty much the last thirty or so years of computer history.

      At one spot in the store, I stopped and picked up a device I hadn't seen for years, and thankfully haven't had to touch one for much longer than that. And then realized that many of my younger networking peers wouldn't have a clue what the heck it was.

      What was it you ask? A thicknet (10Base5) ethernet transciever.

    3. Re:I'm OK with ethernet in cars by swillden · · Score: 1

      I've had my moments of frustration with making Ethernet cables, but like most anything else it's a question of practice. At first, only about one end in five worked, then I got to where I could make them quickly and correctly... but then I went a couple of years without doing it and struggled the next time I had to make a few. At this point, having done it on and off for enough years, I know that the first cable or two I make has to be done slowly and carefully, double-checking everything, but it'll work and the next few will be quick and easy.

      Here in a few months I'm going to wire my new house myself, so I expect to get extremely good at it. The house is new construction so I'm going to take the opportunity while the walls are open to wire the whole thing very heavily with Cat 6 -- by "heavily" I mean, an end point every eight feet in every wall in the house, with all lines down to a rack in the basement. I'm debating putting conduit everywhere, too for future-proofing, but I honestly think that Cat 6 is probably future-proof enough. It'll enable at least 10 Gigabit connections, perhaps faster, and the future seems to be going mostly wireless anyway.

      --
      Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
    4. Re:I'm OK with ethernet in cars by afgun · · Score: 1

      Do the same thing with power plugs; you'll thank yourself later.

    5. Re:I'm OK with ethernet in cars by swillden · · Score: 1

      I'm going to put outlets every six feet along every wall, plus outlets wherever it seems like I even might someday want one. I did that in a previous remodel, and it was nice.

      --
      Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
  36. Aircraft by dbIII · · Score: 1

    Doesn't the Airbus A380 have a pile of ethernet cables running all over the thing?

    1. Re:Aircraft by dizdar · · Score: 1

      Yes ... but it's "avionics grade" (ie. deterministic).

      AFDX

  37. Ethernet for cars? by Anonymous Coward · · Score: 0

    Guess if you are going to run cables all over the fucking roads, you might as well make it POE for all those electric "cars".

  38. Re:Your next car will be a bikeTesla. by Firethorn · · Score: 1

    EVs are very much upcoming. The only thing holding Tesla back from making more cars is that they're already consuming something like half the LiIon cells produced on the planet.

    --
    I don't read AC A human right
  39. Re:Don't quit your day job by fwc · · Score: 1

    Correct me if I'm wrong, but aren't we all hybrid descendants of the Cyclons and humans which fled the Twelve Colonies?

  40. Re:Don't quit your day job by Anonymous Coward · · Score: 0

    Yeah, everyone knows the Cylons are going to show up any day now.

    I fear the Cylons far less than Skynet.

  41. New kind of Ethernet for cars by Anonymous Coward · · Score: 0

    The article mentions a new kind of Ethernet will be used, different from the well known CAT cables and 8P8C connectors.

    It seems it will be Gigabit Ethernet over a single(!) twisted pair. Full duplex ? Nice.
    It is called RTPGE (Reduced Twisted Pair Gigabit Ethernet): http://www.ieee802.org/3/bp/index.html

    Also "PoE" over RTPGE is in the works: http://www.ieee802.org/3/bu/index.html

    I can imagine that this has the potention to not only be used in cars, but also replace many proprietary control busses currently in use.

  42. Good Call by aaronb1138 · · Score: 1

    I for one, was quite terrified after looking over CAN-BUS and similar post ODBC-II standards that the shitheads who design car electronics would start using modified versions of USB 2/3 for automotive communication.

    Seriously look over any automotive electronics spec and you can smell the steaming pile of, "I want to be guaranteed easy employment another 10 years" shit from a mile away. I wish engineering projects included more ethics audits that resulting in contractors and employees getting fired, black balled, and sued.

  43. individual bus for ECU and Entertainment by thygate · · Score: 1

    Is this being proposed as a replacement for the engine-control bus, like the CAN-bus ? Do we really want (relatively high-bandwidth) entertainment shizzle to be on the same bus/network as the as all the engine sensors, actuators and controllers etc. ? Fine if it's just some control signals, like 'volume up', 'play', 'next', 'accept incoming call' .. whatever. But wouldn't it make much more sense to use a separate good-old ethernet cable, or even WiFi to for the actual media devices and the internet. If it's all in the same network, any user-hot-pluggable device on said network is physically able to do crazy stuff, like 'deploy airbags', or 'unlock', 'engine start', etc.. Let's not even think about connecting this to the internet.

  44. How is this big news? by TomGreenhaw · · Score: 1

    Ethernet is already in widespread use in vehicles. The Tesla already uses Ethernet for almost everything mentioned in the article including real time connectivity to the internet for turn by turn directions, streaming music, diagnostics and more. It also uses WiFi for things like tire pressure sensors. Tesla is just one example; many luxury vehicles use WiFi for their tire pressure sensors. Yet another standard for automotive cabling and the use of existing protocols is common sense and hardly a major breakthrough.

    --
    Greed is the root of all evil.
  45. Hundred Meter Tether by sudon't · · Score: 1

    So wait, my car will be on a hundred meter tether to my router? [rimshot]

    --
    -- sudon't

    Air-ride Equipped

  46. Re:Don't quit your day job by roc97007 · · Score: 1

    Correct me if I'm wrong, but aren't we all hybrid descendants of the Cyclons and humans which fled the Twelve Colonies?

    Dunno. I stopped watching partway into the third season.

    --
    Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
  47. a car's head unit by Anonymous Coward · · Score: 0

    so i'm guessing truck drivers have had this for years?

  48. Anyone RTFA? by hoboroadie · · Score: 1

    I hope they specified Cat5e STP.

    --
    They feared that it could be used to suppress protest or support unpopular rule.
  49. IP? Ugh. by afgun · · Score: 1

    Yeah, sorry, but I don't want a protocol in my car that's subject to script kiddies DoSing it. I'm sure they'll get to CAN and everything else soon, but at least for now you can't just go to a warez site and take down a car. I think the crack about Cylons was spot on.

  50. This is news? by Anonymous Coward · · Score: 0

    I was told that all the components in my 2005 Chevy Trailblazer (the model came out in 2002) - instruments, mirrors, seats, etc., are connected via Ethernet. Not 1-gig Ethernet, but who needs 1-gig Ethernet to change the station on the radio?