Robotic Telesurgery by Remote Surgeons 134
Roland Piquepaille writes "In a few years, telesurgery performed by multi-armed robots remotely controlled by real surgeons located hundreds or thousands of kilometers away will become commonplace. Today, Canadian doctors from the Centre for Minimal Access Surgery (CMAS) are developing the technology for NASA. Their goal is to build a portable robotic unit that would be used in space missions, war zones and remote areas within five years. So far, the experiments already done in Canada and for NASA are extremely encouraging. But read more for additional details and pictures of a real surgeon controlling such a robot."
Re:Don't try that on the Internet until ... (Score:3, Interesting)
But it seems obvious to me, at least, that this would only be usable with dedicated links, and not over the internet, which is what a lot of others seem to be saying here. I doubt very much that any of the engineers involved have even considered using the public internet (at least I hope they haven't!
SB
Awake during surgery?? (Score:3, Interesting)
Re:Don't try that on the Internet until ... (Score:3, Interesting)
Only if they're a TDM point-to-point or switched connection, or a virtual one using something like ATM. In which case it's not the Internet.
It's connection-oriented, not packet-oriented (even if it's packets being carried,and they're being switched one at a time.) Connection-oriented networks give you a fixed(-maximum)-bandwidth connection with guaranteed delivery and guaranteed limits on latency and jitter.
Packet switched networks can emulate this, but ONLY with QoS guarantees - guarantees on delivery, latency, and latency variation. That means some packets take precedence over others. They go to the head of the line. When there are too many packets these privileged ones get forwarded while others are dropped. And so on.
This is NOT something you can build out of best-effort delivery and retransmission, as you build reliable (but variable speed and latency) connections out of unreliable ones using protocols like TCP.
Re:Don't try that on the Internet until ... (Score:3, Interesting)
A QoS-enabled Internet would work just fine. And it's coming - unless misguided "Net Neutrality" laws throw the baby out with the bathwater.
There are two aspects to "neutrality" and tiered bandwidth.
- One is using it to distinguish services that need different levels of support - giving them what they need (and perhaps charging extra if appropriate), but treating all services requiring a particular QoS level (including your own) equally.
- The other is using the tools to implement anticompetive practices, such as penalizing your competitors' packets or charging some customers extra just because they have deep pockets to be picked.
The first lets a network provider combine guaranteed-QoS and best-effort traffic on a single network and give the best-effort traffic all the remaining bandwidth once the high-QoS stuff is serviced. This is massively cheaper than either of the alternatives: Separate networks, or a permanent bandwidth split on a single one, with the high-QoS partition large enough to handle the maximum load and its unused bandwidth left idle. This saving ends up going mainly to the network users, in the form of lower rates.
The second is an anti-competitive practice and a worthy target of suppression - by "the invisible hand" if possible, perhaps by law if not.
Comment removed (Score:2, Interesting)