SRAM's wireless shifting—a saboteur's dream?

don't worry, SRAM will use the feedback of early adopters to release a functional second generation. All at premium prices, of course.

My favourite piece of trivia is this: the USS Phoenix was one of the few US ships that survived the Japanese attack on Pearl Harbour with minor damage. It went on until 1982 when it was sunk by the British during the Falklands war.

The answer to the question in your headline then is "No".

Longest post on road cc ever?

I don't have direct experience of 802.15.4, but I am fairly deep into networking (particularly, analysing how network protocols can work and fail).

Note that 802.15.4 specifies the radio (PHY) and media access control/arbitration (MAC) layers. I.e. it specifies everything below the application layer. It can be downloaded at:

http://standards.ieee.org/getieee802/download/802.15.4-2006.pdf
http://standards.ieee.org/getieee802/download/802.15.4-2011.pdf

The patent says they're using the "unlicensed" 2.4 to 2.5GHz band specified by the ITU. This is carved up into 16 relatively narrowband channels. This radio band is also used by a very wide number of other devices, including 802.11 wifi, DECT phones, garage door openers, etc.

The 16 channels does *not* mean it's limited to 16 devices using it. Each of the 16 channels can be used by multiple devices, which have to co-ordinate when they transmit using the specified "MAC" protocol. 802.15.4 is defined to use a CSMA-CA algorithm. This is fairly simple, however performance degrades fairly quickly as more devices use the same channel - similar to when your neighbours' wifi is on the same channel as yours, and/or as you have more and more devices using your wifi (and your 802.11 wifi can use more slightly more sophisticated MAC protocols).

The SRAM patent has a lot of talk about transmitting on a channel, listening for noise, waiting, etc. I suspect though they're just referring to the standard 802.15.4 protocols there - that SRAM didn't invent any new PHY/MAC stuff, but are just using standard implementations. If so, I think it's a bit sneaky of them to try claim this stuff in their patent, but IANAL.

They do seem to have added their own "turn the chip/radio on and off at intervals" protocol on top of that though to save power. These kinds of things can be tricky to get right though. Protocols can work in controlled settings, but then fall apart in certain other conditions that are only experienced in the real-world.

The SRAM patent seems to suggest that channel selection is done manually, by pressing a button on the dérailleur to make it go into a pairing mode where it scans all the channels, while the operator sets the channel on the shifter using some appropriate twiddling after which it transmits. The dérailleur then picks up the channel the shifter sends on.

Finally, the patent suggests the shifter sends fairly simple "Shift up" or "shift down" commands. The patent then has to discuss how the problem of duplicate or lost commands are dealt with. Commands can be lost because of either radio noise, or because of the aforementioned "turn the radio off for intervals" power-saving protocol. To compensate for this, the shifter also sends a count, e.g. "Shift up 3" to indicate this is the 3rd up shift in a row. I would presume the shifter resets this count when an opposite shift is made. The patent as written doesn't actually make sense to me, it says the receiver increments the count, but I think they mean the shifter does - I don't understand how it could work otherwise, unless the dérailleur acknowledged shift counts, but the patent doesn't say that this happens.

When the dérailleur receives a message where the count is higher than the previous message it received, then it knows it missed a message and can compensate by shifting more, to match the number of missed shifts. This means that lost commands from the shifters in noisy RF conditions must definitely be an issue though.

It seems like their protocol should generally work, however I can see some problems if sufficient messages are lost that the rider notices and compensates themselves by pressing the lever. E.g.:

1. Dérailleur last got "shift down, x-1"

1. Rider wants to shift down again, clicks

2. Shifter sends "Shift down, x", and does so at intervals for long enough that the dérailleur must have had an awake period, to have had a chance to receive - e.g. 100ms is suggested in the patent, but that might not what the released systems use.

3. The dérailleur doesn't get the message, because of high radio noise.

4. 1 or 2 seconds later the rider notices the missed shift, and moves the shifter again.

5. Repeat steps 2 to 4 another n times, where n is 0 or higher, depending on how long the radio noise lasts. Each time the shifter is sending an incremented x.

6. Finally the dérailleur receives the message, and suddenly shifts down x+n times.

That wouldn't be fun.

Another issue is that SRAM don't have separate shifters for front and rear dérailleurs. Instead, you have to tap both shifters together. To make this work, the patent says the shifters /actually/ send "Shift up/down pressed" and "shift up/down released" messages. The front and rear dérailleurs then have to work out from the /combination/ of messages whether it was intended for them.

I.e. receiving either:

1. "shift up pressed"

2. "shift up released"

or else:

1. "shift down pressed"

2. "shift down released"

Means it's for the rear dérailleur.

However, if the following is received (where the order of 1a and 1b to each other don't matter, nor 2a to 2b, so long as the 2s are received after the 1s):

1a. "shift up pressed"
1b. "shift down pressed"

2a. "shift up released"
2b. "shift down released"

Then it means it's meant for the front dérailleur, and the FD will change ring. (Clearly, this can't work at all with a triple). However, that means that if either:

c: message 1a is lost, and then message 2b is received

d: message 1b is lost, and then message 2a is received

at the rear dérailleur, then the rear dérailleur will change. It will change down in the case of c, and it will change up in the case of d. Note that it is possible for the rear dérailleur to miss the message, but the FD to receive - in which case *both* will change.

Further, this means this protocol can not intentionally shift both front and rear at the same time (I do this sometimes, e.g. big to little on the front, while shifting to a smaller cog on the back). Of course, both-shifting generally isn't recommended. However, this protocol in the patent can *unintentionally* shift both.

It will be interesting to see how often the problem of "over shifts" (shifting too many because of lost shift) occurs on the rear, and how often the "unintentional both shift" occurs, due to RF noise.

NB: It is possible that SRAM further developed their shifting protocols since the patent applications, and so the might have already addressed these issues.

Simply using AES does not necessarily mean it will be harder to hack. It's all down to cryptographic key management - i.e. how complex/long the key is (i.e. to prevent brute force attack), how well the key is protected/stored on the device and whether or not each groupset will have its own unique key. I would hope the keys are unique, because if you crack one gruppo, well ....

I see a lot of crappy implementations of cryptography on mobile apps and I can find AES keys to encrypted databases stored on the impactded device(s) within minutes. None of this is down to any deficiencies in AES itself as a technology, moreso because of security being an afterthought and corners being cut in the interests of costs and development time.

It would be an interesting test to see what is possible from passively scanning the wireless communications. But .... the attacker would always have to be in close physical proximity to the groupo to be able to do that.

However, I'm inclined to be optimistic and hope that SRAM have taken the necessary steps to adequately test the security of their new groupo. Let's see.

Getting off my hobby horse about cryptography, another interesting thing that nobody has talked about is whether or not the new technology can integrate with head units. For instance, as well as recording speed, cadence, heart rate, elevation, etc, etc, why not add gear usage to that as well? Could it even enable head units to display gear information too (for those that can't be bothered to look down at their chainset and cassette). Interesting times!

p.s. 'spotty hackers'? Makes me laugh when I see this stereotype.

....multiple shift buttons being a feature of the ancient mavic zap mektronic wireless electric Gruppo.

Yes, multiple shifters is the obvious way of progress on this product. I'm also waiting for splitters for hydraulic hoses which would allow riders to use extra brake levers at their TT extensions or install hydraulic CX-style levers. SRAM have recently started shipping the new generation of Hydro-R brakes/levers so who knows... That would be pretty nice!

Should be easy to attach more shifters as well

I'd consider it after a couple of recalls have sorted out the early bugs. BUT, still three (four? both shifters?) batteries to charge.

Great photo of Hedy here are a few quotes from her

"Any girl can be glamorous. All you have to do is stand still and look stupid."

"If you use your imagination, you can look at any actress and see her nude. I hope to make you use your imagination."

"I must quit marrying men who feel inferior to me. Somewhere there must be a man who could be my husband and not feel inferior. I need a superior inferior man."

"The world isn't getting any easier. With all these new inventions I believe that people are hurried more and pushed more... The hurried way is not the right way; you need time for everything - time to work, time to play, time to rest."

"Perhaps my problem in marriage - and it is the problem of many women - was to want both intimacy and independence. It is a difficult line to walk, yet both needs are important to a marriage."

"Hope and curiosity about the future seemed better than guarantees. That's the way I was. The unknown was always so attractive to me... and still is."

With regards to these wireless gears they are just the natural progression to automatic gears in the future. I foresee a time where the gear changes to your bodies power output, lactate levels and pain threshold etc.

Does "16 channel" limit the number of devices able to operate in one location? If so, a peloton-full would be disastrous/comedy gold.