shimano auto shift

[Roger Meriman]

Shimano has auto shifting/semi auto shifting technology which looked
interesting on Pinkbike also has a nice car!

https://youtu.be/2WF7UFULQ6w

Shimano page.

https://bike.shimano.com/en-EU/technologies/component/details/steps/auto-shift.html

Definitely interesting technology, I’d certainly be interested in trying it
possibly a solution in search of a problem, though I’m not it’s target
audience ie I’m fairly technical on the MTB plus it’s a fairly mature bike!

Roger Merriman

[Lou Holtman]

On Friday, July 21, 2023 at 8:14:24 PM UTC+2, Roger Meriman wrote:
> Shimano has auto shifting/semi auto shifting technology which looked
> interesting on Pinkbike also has a nice car!
>
> https://youtu.be/2WF7UFULQ6w
>

You need an E atb?

Lou

[Roger Meriman]

Would seem so yes.

Roger Merriman

[Lou Holtman]

Hmm they already ‘cheating’ and now they are to lazy/incompetent to shift? Shifting at the right time is part of the fun/skill. Setting the shifting profile in an app? Do you have to do that again when the terrain changes?

Lou

[Frank Krygowski]

I couldn't make it through the video. The ratio of actual information to
screen time was far too low.

--
- Frank Krygowski

[Andre Jute]

> Roger Merman
>
Pft! I had Shimano full auto shifting on my Trek Navigator L700 "Smover" twenty years ago. See
http://coolmainpress.com/BICYCLINGsmover.html
>
To answer Lou's sardonic question, sure, I found the auto box very useful. But then I ride in the countryside looking around me, talking with companions, enjoying myself rather than hunched over drop bars like a gorilla in too small a cage. Excluding Porsche which didn't come with an outbox until I was already middle-aged. the last time I had a manual box in anything except a racing car, it was in an MG TC, a miserable little piece of shit. My fave car, before I switched to Citroen's utterly exceptional SM and then to the Bentley Turbo via a stint in a Jensen Interceptor with a big block Chrysler engine, for crossing Europe overnight from London to Nardo in the boot of Italy was a 22-foot long 7 litre Ford with suspension and brakes I'd breathed on. I just chuckle at middle-aged men who buy a stick shift because they think it is faster; soon it stands unloved in their garage, just like the backbreaking bikes they thought would make them look younger, and a used car dealer makes killing buying it cheap and selling it dear.
>
It looks to me like Shimano cut the eight gears on the Nexus Premium 8-speed HGB I have down to 5 for electric mountainbikes (that's an oxymoron!). I wrote in an article on another forum -- where work actually gets done without inference from scum, unlike here on RBT -- that an e-bike with the motor and battery scaled to the weight being carried, and proper software of course, does not need even the seven gears in the upper half of Rohloff's Speed 14 HGB, though the overdrive gears 12-14 are very nice to have. Of course, unless the motor is wimpish and the battery cramped, you need the sturdiness of the Rohloff because a good electric motor will rip Shimano's Nexus apart.
>
The example I used for the demonstration was my Bafang BB central motor installation, which in the software supplied with it has 9 electronic gears and the ability to reduce the number of these gears without reducing the range. On my Utopia Kranich I have the electric motor's gear number set to 5, and I use the top seven gears of a well-broken in Rohloff to fine-tune the speed to my cadence, when I can be bothered or in bigger jumps if my heart rate -- in plain view on an iPhone on the handlebars -- should rise above the permitted limit.
>
It looks to me like Shimano followed the same logic, possibly to leave space and weight allocation to make the 'box stronger than either the 4-speed or 8-speed Premium Nexus. And make no bones about it, the Nexus boxes need strengthening to guarantee a decent service life when coupled to an electric motor, considering that with just the power of my legs I managed to wreck two with less than 5000km between them. I'm slightly surprised that Shimano didn't switch over to CVT as in the NuVinci but I suppose CVT is hedged-about by patents and Shimano likes to do their own innovation. The NuVinci has a good reputation for longevity and it already offers an automatic "gearbox".
>
Also, should give Shimano credit for bringing back and persevering with a good idea even in the face of setbacks, and as always for offering truly complete high value group sets.
>
Andre Jute
Now let us praise famous men. -- Ecclesiastes
>

[Andre Jute]

The aim of the software which signals the gear change to a simple small motor on the end of the axle is to match the gear and thus the effort of the driver to the terrain. Don't think of a variable speed motor. Instead start with the idea of constant output from the drive (electric motor, legs) and the terrain as the variable input. If you could learn to adapt the derailleur gears on your bike perfectly to the terrain, you cadence would never be broken. That's what an autobox on a bike does. Everything else is a distraction.
>
Andre Jute
It takes a while to grasp it, but I think it is likely that the NuVinci autobox will eventually be known as the optimum of bike autoboxes, precisely because it is tied a CVT. You heard it here first.
>

[Catrike Rider]

On Fri, 21 Jul 2023 15:49:38 -0700 (PDT), Andre Jute
<fiul...@yahoo.com> wrote:

>On Friday, July 21, 2023 at 8:29:11?PM UTC+1, Lou Holtman wrote:

>> On Friday, July 21, 2023 at 9:15:35?PM UTC+2, Roger Meriman wrote:
>> > Lou Holtman <lou.h...@gmail.com> wrote:

>> > > On Friday, July 21, 2023 at 8:14:24?PM UTC+2, Roger Meriman wrote:
>> > >> Shimano has auto shifting/semi auto shifting technology which looked
>> > >> interesting on Pinkbike also has a nice car!
>> > >>
>> > >> https://youtu.be/2WF7UFULQ6w
>> > >>
>> > >
>> > > You need an E atb?
>> > >
>> > > Lou
>> > >
>> > >
>> > >
>> > Would seem so yes.
>> >
>> > Roger Merriman
>> Hmm they already ‘cheating’ and now they are to lazy/incompetent to shift? Shifting at the right time is part of the fun/skill. Setting the shifting profile in an app? Do you have to do that again when the terrain changes?
>>
>> Lou
>>
>The aim of the software which signals the gear change to a simple small motor on the end of the axle is to match the gear and thus the effort of the driver to the terrain. Don't think of a variable speed motor. Instead start with the idea of constant output from the drive (electric motor, legs) and the terrain as the variable input. If you could learn to adapt the derailleur gears on your bike perfectly to the terrain, you cadence would never be broken. That's what an autobox on a bike does. Everything else is a distraction.
>>
>Andre Jute
>It takes a while to grasp it, but I think it is likely that the NuVinci autobox will eventually be known as the optimum of bike autoboxes, precisely because it is tied a CVT. You heard it here first.
>>

Doesn't work for people like me who like to vary their cadence.

[Roger Meriman]

She does reference that, note she is good Enduro racer ie that for that
sort of use probably will want full control though I can certainly see
advantages in a bike that is more tolerant of poor shift timings, ie a bike
that could shift while freewheeling potentially could be useful?

But I’ve not ridden it nor likely to as it’s not my type of riding my MTB
is much more XC like, ie I like the physical challenge of hills both in
terms of fitness and technique, I’m less concerned with the downs.

My friends have E MTB and work well for them but the advantages for me
considering where I like and all that just isn’t worthwhile.

Roger Merriman

[Roger Meriman]

That’s why I linked the document, but she does a fairly good of explaining
ie it’s a part of the Shimano MTB Di2 + E bike system mainly that it can
shift while your freewheeling as you might on technical DH stuff, and the
motor can slowly spin the drive chain to enable this.

And as ever you can have full control or auto or somewhere in between.

I don’t use EMTB nor likely to unless life changes drastically. But as
technology it seems interesting.

Roger Merriman

[Andre Jute]

It adapts to whatever your cadence is in the moment, if that's what you want. In any event, you can tell the computer specifically what your next cadence will be, and it adapts instantly, if you feel you don't want a computer with the intelligence of a cat making decisions for you. In fact I found Shimano's programming more than adequate, so much so, in fact, that I set the computer up before my first ride and never touched it again; it took me only fractionally longer to arrive at a modus operandi with the control computer of my electric bike, which is clearly much more sophisticated.These things just work for me, perhaps in recognition of the fact that I've been in computers since they had glowing valves, what Americans call thermionic tubes, and I am under no delusion that artificial intelligence will any time soon be able to do what I do, which is create things, conceive things which nobody else has done before. Never forget: Garbage in, garbage out.
>
Andre Jute
Artificial intelligence is one of the most insidiously pernicious oxymorons of our time.
>

[Catrike Rider]

On Sat, 22 Jul 2023 20:32:42 -0700 (PDT), Andre Jute
<fiul...@yahoo.com> wrote:

>On Saturday, July 22, 2023 at 9:30:25?AM UTC+1, Catrike Rider wrote:
>> On Fri, 21 Jul 2023 15:49:38 -0700 (PDT), Andre Jute
>> <fiul...@yahoo.com> wrote:
>>
>> >On Friday, July 21, 2023 at 8:29:11?PM UTC+1, Lou Holtman wrote:
>> >> On Friday, July 21, 2023 at 9:15:35?PM UTC+2, Roger Meriman wrote:
>> >> > Lou Holtman <lou.h...@gmail.com> wrote:
>> >> > > On Friday, July 21, 2023 at 8:14:24?PM UTC+2, Roger Meriman wrote:
>> >> > >> Shimano has auto shifting/semi auto shifting technology which looked
>> >> > >> interesting on Pinkbike also has a nice car!
>> >> > >>
>> >> > >> https://youtu.be/2WF7UFULQ6w
>> >> > >>
>> >> > >
>> >> > > You need an E atb?
>> >> > >
>> >> > > Lou
>> >> > >
>> >> > >
>> >> > >
>> >> > Would seem so yes.
>> >> >
>> >> > Roger Merriman
>> >> Hmm they already ‘cheating’ and now they are to lazy/incompetent to shift? Shifting at the right time is part of the fun/skill. Setting the shifting profile in an app? Do you have to do that again when the terrain changes?
>> >>
>> >> Lou
>> >>
>> >The aim of the software which signals the gear change to a simple small motor on the end of the axle is to match the gear and thus the effort of the driver to the terrain. Don't think of a variable speed motor. Instead start with the idea of constant output from the drive (electric motor, legs) and the terrain as the variable input. If you could learn to adapt the derailleur gears on your bike perfectly to the terrain, you cadence would never be broken. That's what an autobox on a bike does. Everything else is a distraction.
>> >>
>> >Andre Jute
>> >It takes a while to grasp it, but I think it is likely that the NuVinci autobox will eventually be known as the optimum of bike autoboxes, precisely because it is tied a CVT. You heard it here first.
>> >>
>> Doesn't work for people like me who like to vary their cadence.
>>
>It adapts to whatever your cadence is in the moment, if that's what you want. In any event, you can tell the computer specifically what your next cadence will be, and it adapts instantly, if you feel you don't want a computer with the intelligence of a cat making decisions for you.

No, that wouldn't work for me, either. I don't pre-arrange my cadence,
I just do it, and then shift to accomodate it. At any rate, I'll keep
on manually shifting.

[Roger Meriman]

Which you can with the system, looks like a paddle buttons in apparently
free ride mode!

It’s tied into the E bike system so will not be as pure? As a analogue bike
ie a degree of auto system happens anyway, and control of how much help be
that power or gears shifted and so on.

Roger Merriman

[Catrike Rider]

On Sun, 23 Jul 2023 11:23:22 GMT, Roger Meriman <ro...@sarlet.com>
wrote:

>Catrike Rider <sol...@drafting.not> wrote:
>> On Sat, 22 Jul 2023 20:32:42 -0700 (PDT), Andre Jute
>> <fiul...@yahoo.com> wrote:
>>
>>> On Saturday, July 22, 2023 at 9:30:25?AM UTC+1, Catrike Rider wrote:
>>>> On Fri, 21 Jul 2023 15:49:38 -0700 (PDT), Andre Jute
>>>> <fiul...@yahoo.com> wrote:
>>>>
>>>>> On Friday, July 21, 2023 at 8:29:11?PM UTC+1, Lou Holtman wrote:
>>>>>> On Friday, July 21, 2023 at 9:15:35?PM UTC+2, Roger Meriman wrote:
>>>>>>> Lou Holtman <lou.h...@gmail.com> wrote:
>>>>>>>> On Friday, July 21, 2023 at 8:14:24?PM UTC+2, Roger Meriman wrote:
>>>>>>>>> Shimano has auto shifting/semi auto shifting technology which looked
>>>>>>>>> interesting on Pinkbike also has a nice car!
>>>>>>>>>
>>>>>>>>> https://youtu.be/2WF7UFULQ6w
>>>>>>>>>
>>>>>>>>
>>>>>>>> You need an E atb?
>>>>>>>>
>>>>>>>> Lou
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>> Would seem so yes.
>>>>>>>
>>>>>>> Roger Merriman

>>>>>> Hmm they already ?cheating? and now they are to lazy/incompetent to

Interesting systems, but there might even be something better in 15
years or so, when I might possibly look at E-bikes ;) I'm not there
yet.

[Roger Meriman]

I’m sure it will progress quite quickly to be honest!

I’m unconvinced that it’s what I want but as technology sounds interesting
and all that.

Roger Merriman

[Catrike Rider]

On Mon, 24 Jul 2023 10:11:06 GMT, Roger Meriman <ro...@sarlet.com>

>>> It?s tied into the E bike system so will not be as pure? As a analogue bike

>>> ie a degree of auto system happens anyway, and control of how much help be
>>> that power or gears shifted and so on.
>>>
>>> Roger Merriman
>>
>> Interesting systems, but there might even be something better in 15
>> years or so, when I might possibly look at E-bikes ;) I'm not there
>> yet.
>>
>I’m sure it will progress quite quickly to be honest!
>
>I’m unconvinced that it’s what I want but as technology sounds interesting
>and all that.
>
>Roger Merriman

There are things about AI that concern me. It's very different from
biological intelligence in that biological intelligence dies when the
host/body dies.

Theoretically, AI does not necessarily do that. It can transfer and/or
be cloned. It can repair itself and it's host/body.

[Roger Meriman]

Don’t believe is any Ai in that technology nor likely to.

Any more than my Gravel/MTB light that will learn to a degree your riding
style with the various giro’s etc in terms of light brightness, ie will
auto dim if stopped and power up, once you move, kicking further up as the
riding gets faster or more intense ie less oomph on fire roads vs trails.

Roger Merriman

[Catrike Rider]

On Mon, 24 Jul 2023 12:16:23 GMT, Roger Meriman <ro...@sarlet.com>

>>> I?m sure it will progress quite quickly to be honest!
>>>
>>> I?m unconvinced that it?s what I want but as technology sounds interesting

>>> and all that.
>>>
>>> Roger Merriman
>>
>> There are things about AI that concern me. It's very different from
>> biological intelligence in that biological intelligence dies when the
>> host/body dies.
>>
>> Theoretically, AI does not necessarily do that. It can transfer and/or
>> be cloned. It can repair itself and it's host/body.
>>
>Don’t believe is any Ai in that technology nor likely to.
>
>Any more than my Gravel/MTB light that will learn to a degree your riding
>style with the various giro’s etc in terms of light brightness, ie will
>auto dim if stopped and power up, once you move, kicking further up as the
>riding gets faster or more intense ie less oomph on fire roads vs trails.

All this *IS* AI... Simple AI, but AI, just the same.

>Roger Merriman

[AMuzi]

A recent brief filed by an actual licensed practicing
attorney was written with such a system:

https://interestingengineering.com/culture/lawyer-used-chatgpt-to-write-legal-brief-bogus

Which cited a fictitious judge's decision in a fictitious
jurisdiction on a fictitious case. But in nicely crafted prose.

This is not a practice we ought to embrace and holds great
risk once better established.

--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971

[Tom Kunich]

The question arises - what was the reaction of the Bar Association?

[AMuzi]

On 7/24/2023 9:55 AM, Tom Kunich wrote:

>>>> I’m sure it will progress quite quickly to be honest!
>>>>
>>>> I’m unconvinced that it’s what I want but as technology sounds interesting

>>>> and all that.
>>>>
>>>> Roger Merriman
>>>
>>> There are things about AI that concern me. It's very different from
>>> biological intelligence in that biological intelligence dies when the
>>> host/body dies.
>>>
>>> Theoretically, AI does not necessarily do that. It can transfer and/or
>>> be cloned. It can repair itself and it's host/body.
>>>
>> A recent brief filed by an actual licensed practicing
>> attorney was written with such a system:
>>
>> https://interestingengineering.com/culture/lawyer-used-chatgpt-to-write-legal-brief-bogus
>>
>> Which cited a fictitious judge's decision in a fictitious
>> jurisdiction on a fictitious case. But in nicely crafted prose.
>>
>> This is not a practice we ought to embrace and holds great
>> risk once better established.
>>

> The question arises - what was the reaction of the Bar Association?
>

Well he did express his remorse in being caught to the Court:

https://nypost.com/2023/06/08/humiliated-lawyer-apologizes-over-chatgpt-court-flub/

Didn't find a NY Bar Association complaint.

[Tom Kunich]

When the universities started talking about AI I was one of the first in private business to program AI into medical instruments. So I have a very clear picture of its limitations. What this actually sounds like to me is that they have developed an AI for writing fiction and it cannot be used outside of that function. This does put an extra burden on teachers to actually read and grade homework on its correctness rather than wordcount, but AI despite words to the reverse, is no threat to people.

[Lou Holtman]

Except if a lot of people believe the nonsense that can come out of an algorithm called AI and act accordingly. It is all about what dataset the algorithm uses.

Lou

[Roger Meriman]

It’s more of a calculation on fairly simple parameters, my car’s automatic
I think “learns” but this is just a very simple calculation.

Same as auto focus/exposure/light balance it’s sophisticated but quite
limited no intelligence.

And to be honest most AI etc images are fairly easy to spot. Often hair and
hands.

And well if something is too good to be true, worth pondering if it’s true
or not ie photos of half naked famous women particularly politicians seen
one of alexandria ocasio-cortez who I’m vaguely aware of.

Where the hair texture was wrong, and having a scan down images she doesn’t
do big cleavages so this image on both raised red flags as to its reality!

Roger Merriman

[Tom Kunich]

I agree with you. But I am afraid that too many teachers aren't themselves learned enough in the subjects they are teaching to correct outright BS. I believe that Frank is a real mechanical engineer. But why does the college he taught at hesitate to rate him so?

[Catrike Rider]

On Mon, 24 Jul 2023 17:31:38 GMT, Roger Meriman <ro...@sarlet.com>

>>> Don?t believe is any Ai in that technology nor likely to.

>>>
>>> Any more than my Gravel/MTB light that will learn to a degree your riding

>>> style with the various giro?s etc in terms of light brightness, ie will

>>> auto dim if stopped and power up, once you move, kicking further up as the
>>> riding gets faster or more intense ie less oomph on fire roads vs trails.
>>
>> All this *IS* AI... Simple AI, but AI, just the same.
>>
>>> Roger Merriman
>>
>It’s more of a calculation on fairly simple parameters, my car’s automatic
>I think “learns” but this is just a very simple calculation.
>
>Same as auto focus/exposure/light balance it’s sophisticated but quite
>limited no intelligence.
>
>And to be honest most AI etc images are fairly easy to spot. Often hair and
>hands.
>
>And well if something is too good to be true, worth pondering if it’s true
>or not ie photos of half naked famous women particularly politicians seen
>one of alexandria ocasio-cortez who I’m vaguely aware of.
>
>Where the hair texture was wrong, and having a scan down images she doesn’t
>do big cleavages so this image on both raised red flags as to its reality!
>
>Roger Merriman

My definition of AI is when a device alters it's behavior/output
without any manual input from a human. The camera on my garage and my
doorbell security systems are not AI because they turn on when
anything moves in their field of vision. They would, in my opinion, be
AI if they were triggered only by a human. It would have to make
assumptions and decisions about what is the difference between a human
and everything else it might sense.

[Roger Meriman]

That’s a rather broad definition! Which would include most washing machines
which use the weight of clothes to judge the length of a wash, that’s why
the estimated time is that a estimate but it’s fairly simple.

Same with auto gearing systems they are reacting to essentially pre defined
conditions, ie cadence/torque and so on.

As ever a good rider/driver is better a mediocre one less so!

Roger Merriman

[Catrike Rider]

On Mon, 24 Jul 2023 17:58:00 GMT, Roger Meriman <ro...@sarlet.com>

>>> It?s more of a calculation on fairly simple parameters, my car?s automatic
>>> I think ?learns? but this is just a very simple calculation.
>>>
>>> Same as auto focus/exposure/light balance it?s sophisticated but quite

>>> limited no intelligence.
>>>
>>> And to be honest most AI etc images are fairly easy to spot. Often hair and
>>> hands.
>>>

>>> And well if something is too good to be true, worth pondering if it?s true

>>> or not ie photos of half naked famous women particularly politicians seen

>>> one of alexandria ocasio-cortez who I?m vaguely aware of.
>>>
>>> Where the hair texture was wrong, and having a scan down images she doesn?t

>>> do big cleavages so this image on both raised red flags as to its reality!
>>>
>>> Roger Merriman
>>
>> My definition of AI is when a device alters it's behavior/output
>> without any manual input from a human. The camera on my garage and my
>> doorbell security systems are not AI because they turn on when
>> anything moves in their field of vision. They would, in my opinion, be
>> AI if they were triggered only by a human. It would have to make
>> assumptions and decisions about what is the difference between a human
>> and everything else it might sense.
>>
>That’s a rather broad definition! Which would include most washing machines
>which use the weight of clothes to judge the length of a wash, that’s why
>the estimated time is that a estimate but it’s fairly simple.

If it always made the length the same for the same weight it would not
fit my definition. If it also changed the cycle according to it's
assumptions about the fabric in the load, that might fit my
definition, depending on how it detirmined the fabric.

[Andre Jute]

Exactemant! AI would be the ability to distinguish between a shark and an attorney on your porch. -- AJ
>

[John B.]

On Mon, 24 Jul 2023 08:52:29 -0400, Catrike Rider

Way back in 1950 Alan Turing devised the "Turing Test".

"A human evaluator would judge natural language conversations between
a human and a machine designed to generate human-like responses. The
evaluator would be aware that one of the two partners in conversation
was a machine, and all participants would be separated from one
another.
The conversation would be limited to a text-only channel, such as a
computer keyboard and screen, so the result would not depend on the
machine's ability to render words as speech.
If the evaluator could not reliably tell the machine from the human,
the machine would be said to have passed the test.
The test results would not depend on the machine's ability to give
correct answers to questions, only on how closely its answers
resembled those a human would give."

He predicted that programs would be capable of fooling judges 30
percent of the time by the year 2000.
--
Cheers,

John B.

[Catrike Rider]

I hope I'm not around to witness when mankind creates an entity that's
not only smarter than us, has a better memory, has better/faster
access to data, can wirelessly transfer itself to another physical
body/host, clone itself simarlarly, and thus survive the destruction
of it's initial body/host....