Does this count as a unicycle?

[esaj]

This is something new, never seen one like this before:

http://www.aliexpress.com/item/Unicycle-electric-scooter-scooter-non-detachable-brake-handle-single-wheel-balance-car-sports-car/32430125153.html

Not sure if I'd like one, the riding picture looks pretty awkward (do you have to bend down to get to the brake lever or what is that?)

[SirGeraint]

Not by my literal definition since it has 2 wheels on it.  Unlike the training wheels on Electric Unicycles it looks like it's supposed to stay.

It isn't self-balancing (not that you have to be to be called a unicycle).  But it's not what we call an Electric Unicycle.

[Mono]

Here is the (somewhat ridiculous) commercial:

and another

and then a thing with the additional wheel behind instead of in the front

If you don't mind, I will avoid to answer your question though.  

[esaj]

Looks like the Hovertrax-clones are now in, there's probably more results with those than actual EUCs when searching for "electric unicycle" in Aliexpress... Found a at least one new shell (but still 350W/132Wh etc.), mostly just the same old generics with a gazillion different brand names.

Edit: Oh right, there's a Firewheel-clone called Macwheel that some sellers have, it looks EXACTLY like a Firewheel, except it says Macwheel on the side, and it's 350W / 174Wh... And it looks like Suoku has stolen Kingsong's shell-design, or the seller is using wrong pictures (which wouldn't be the first time either..)

Oh, here it is:  http://www.suokuwheel.com/suokuwheel/products/2159337.html  except, it's a twin-wheeler (and the Aliexpress-seller IS using Kingsong-pictures for it...)

 

[EUC Extreme]

I do not understand why you keep coming up with these EUC is invented. It is hands-free and everything is easier. and simpler.
If any trouble that keeps stand, forge a good seat. Or even the sofa
But in all of those tunings wonder where is one of the big wheel and small. I do not see any more potential.

[Chuts]

Its an interesting idea this, part bicycle and skateboard.  From what I can tell its not electric powered, so fun on flats and downhill but gonna be a pain to go up hill.

[robca]

Self balancing uses a lot of power, since the powerful motor must keep using lots and lots of torque to counteract any unbalance (and you keep losing momentum). Jason published graphs that show how spiky the power deliver to the motor is (or, better, how much current keeps going in and out of the motor to keep the balance). An identical battery with an identical motor in a two wheeled "device", would have probably at least 30% better range, if not more. Also, EUCs need much bigger and heavier motors (to balance) than would otherwise be required

As a matter of fact, electric skateboards with much smaller batteries can attain the same range as EUCs, even if the investment in electric skateboards these days is much smaller than in EUC (for some reason China doesn't seem to like electric skateboards). Electric skateboard with good range weigh half as much as a EUC with the same range (since they can use much lighter motors)

EUCs are fun, very agile and good at dealing with irregular surfaces, but as a long distance urban transport device, not an efficient design. The power necessary to briefly accelerate and then coast is much smaller than the constant power required to balance and all the constant braking (even with energy recovery, braking briefly to restore balance wastes power, compared to smooth coasting).

So there is value in exploring alternative designs that do no need self balancing and keep as many as the EUC advantages as possible

[Gimlet]

I'm sure I saw those vides a few months ago. Looks fun for a non powered downhill device but having that small wheel out front will seriously limit the size of obstacles it can cope with. I have to wonder why they chose to put the small wheel up front instead of behind as in the more normal penny farthing set up as that would have coped with obstacles more easily.

[Jason McNeil]

@Robca has some good points about eWheels/EUs being less efficient than skateboards & eBikes, but if an eWheel is equipped with >680Wh battery packs, does it really matter? Even cruising along continuously at 28kph, you'll still able to achieve 40km in under one & half hours. This distance/time performance is comparable to popular eBikes, with the advantage of convenience & fun of Wheeling along with a lower total cost of ownership. Comparable Wh/km for electric cars, is about 15-20x more than your typical eWheel (see link below), so personally I don't think the marginal benefits of doubling the efficiency/range are all that important, provided manufacturers can produce a powerful eWheel with the 50-80km range but is also practicable.   
http://my.teslamotors.com/forum/forums/tracking-lowest-kwhkm-whm

I don't believe there are any electric skateboard manufacturers with anywhere near the capacity/range options of the eWheels. 

[Chuts]

@Gimlet Interesting point.  Although I guess there is no reason why you can't ride it backwards.  When I first saw this I thought it was the big wheel that was suppose to be the front as it would make more sense for aerodynamic reasons.

[robca]

@Robca has some good points about eWheels/EUs being less efficient than skateboards & eBikes, but if an eWheel is equipped with >680Wh battery packs, does it really matter? Even cruising along continuously at 28kph, you'll still able to achieve 40km in under one & half hours. This distance/time performance is comparable to popular eBikes, with the advantage of convenience & fun of Wheeling along with a lower total cost of ownership. Comparable Wh/km for electric cars, is about 15-20x more than your typical eWheel (see link below), so personally I don't think the marginal benefits of doubling the efficiency/range are all that important, provided manufacturers can produce a powerful eWheel with the 50-80km range but is also practicable.   
http://my.teslamotors.com/forum/forums/tracking-lowest-kwhkm-whm

I don't believe there are any electric skateboard manufacturers with anywhere near the capacity/range options of the eWheels. 

It matters from a weight and air travel point of view. Today I went to the farmers market. I could have put a 6-7Kg electric skateboard on my back while shopping, and safely made it back and forth. I could not lug a 13-15Kg EUC in the same conditions. Personally I rarely need >15km range, since there are plenty of opportunities to recharge and rarely I need that range. Ability to carry my "vehicle" with me easily, though, is critical. I could easily fly to a new city with an electric skateboard and carry a lot of batteries (<100 Wh), much less so with an EUC (even if hopefully KingSong's new removable pack will help). You seem to be happy to trade off more weight for better distance... me, I'd rather do the opposite ("horses for courses", I guess :)). How many people would really commute 40 Km a day in a open unicycle? compared to the people who would use it to cover long distances during their day, and carry/store the device in between

Electric skateboards can have much bigger ranges than what you see today: most of the existing designs use standard RC LiPos, and you can put as many as you want in parallel: plenty of space on a skateboard deck, or easy to carry on a backpack. An electric skateboard inevitably has better range than a EUC, since it uses a more efficient design (slightly higher friction, due to the 4 wheels). It's just that most skateboard riders don't seem to care: once you are above your daily commute range, no need to carry more weight

EUCs are unbeatable in their ability to negotiate bumps and uneven terrain, though. And more maneuverable in tighter spaces. But one of the heaviest solutions due to the need of powerful motors with a lot of torque

My man point was that, as a transportation device, it's well worth experimenting with new designs that can increase efficiency without requiring huge additional batteries. Since battery technology is the biggest limiting factor (energy density is what it is, and weight/Wh is unlikely to dramatically improve in a short time), the best hope for a viable urban transportation solution is a more efficient design (and fast recharge batteries)

[Mono]

I'm sure I saw those vides a few months ago. Looks fun for a non powered downhill device but having that small wheel out front will seriously limit the size of obstacles it can cope with. I have to wonder why they chose to put the small wheel up front instead of behind as in the more normal penny farthing set up as that would have coped with obstacles more easily.

In my imagination, they are considering the role of the front wheel to be the support during braking. Then it makes perfectly sense to have it in front. Otherwise I agree, any device with a small front wheel is of rather limited practical use beyond toy status. The YikeBike in the third clip I posted has the small wheel in the back (and is electric). 

[Mono]

Self balancing uses a lot of power, since the powerful motor must keep using lots and lots of torque to counteract any unbalance (and you keep losing momentum). Jason published graphs that show how spiky the power deliver to the motor is (or, better, how much current keeps going in and out of the motor to keep the balance). An identical battery with an identical motor in a two wheeled "device", would have probably at least 30% better range, if not more. Also, EUCs need much bigger and heavier motors (to balance) than would otherwise be required

I actually have doubts that balancing uses much energy with a mediocre skilled driver (BTW: do we say driver or rider or...?). I am thinking of an inverted pendulum. It can be balanced with very little effort given we know how to. My random guess would be that additional wheels would be in the same ballpark for their additional necessary energy. 

I wonder: do we have real data on this? How can we even differentiate between "balancing effort" and "moving effort" when measuring the energy flow? I even don't see that there is any balancing effort that needs energy that doesn't give us also forward momentum when we are moving. Keeping constant speed and balancing (forward-backwards) seem to be the very same thing. Keeping constant speed seems however not particularly costly beyond of what needs to be invested in keeping speed anyway.

[Mono]

As a matter of fact, electric skateboards with much smaller batteries can attain the same range as EUCs, even if the investment in electric skateboards these days is much smaller than in EUC (for some reason China doesn't seem to like electric skateboards).

Can you provide comparative numbers, how many km per Wh? 

Electric skateboard with good range weigh half as much as a EUC with the same range (since they can use much lighter motors)

Weight shouldn't make a difference when it comes to energy considerations, as most of the weight is carried by the driver. Weight is a big concern when moving the device around without driving. 

[Mono]

EUCs are unbeatable in their ability to negotiate bumps and uneven terrain, though. And more maneuverable in tighter spaces. But one of the heaviest solutions due to the need of powerful motors with a lot of torque

My man point was that, as a transportation device, it's well worth experimenting with new designs that can increase efficiency without requiring huge additional batteries. Since battery technology is the biggest limiting factor (energy density is what it is, and weight/Wh is unlikely to dramatically improve in a short time), the best hope for a viable urban transportation solution is a more efficient design (and fast recharge batteries)

There are two points to differentiate:

• the motor needs sufficient torque 
• how energy-efficient is the device in terms of km/Wh.

I tend to agree with Jason McNeil, that energy efficiency in itself seems not the main issue: just the battery pack of a short- to medium-range wheel, say 200Wh=15km, weighs far below 2kg. So the design issues seem to be rather the motor and/or the wheel itself.

[esaj]

I actually have doubts that balancing uses much energy with a mediocre skilled driver (BTW: do we say driver or rider or...?). I am thinking of an inverted pendulum. It can be balanced with very little effort given we know how to. My random guess would be that additional wheels would be in the same ballpark for their additional necessary energy. 

I wonder: do we have real data on this? How can we even differentiate between "balancing effort" and "moving effort" when measuring the energy flow? I even don't see that there is any balancing effort that needs energy that doesn't give us also forward momentum when we are moving. Keeping constant speed and balancing (forward-backwards) seem to be the very same thing. Keeping constant speed seems however not particularly costly beyond of what needs to be invested in keeping speed anyway.

I'm not a native english-speaker, but I think it's "rider" (you ride a horse, bicycle, motorcycle, you drive a car). And AFAIK, the balancing of Segway & EUCs are based on the inverted pendulum -equations (see for example here:  http://forum.electricunicycle.org/topic/442-diy-saddled-euc-and-segway-clone-source-code/#comment-4524 )

You can see in @Jason McNeil's Kingsong tests that the power usage is very spiky (this is going uphill at full speed):

I have no other explanation for the spikes except the wheel changing the rotational speed to stay balanced. While the changes may not be that big, they must happen really fast (acceleration/strong deceleration for a split second), and the faster acceleration/deceleration you want, the more power is needed, so balancing might require more power than we think.

At the beginning (0-0.2 on timeline), the wheel looks like it's stationary, then there's a huge spike as it starts going forward, then it steadies around after 0.2 between 100-250 watts. Then it might be stationary again or going steady speed at level ground / very slight downhill without much need for balancing, and then begins the climb around 0.35.  After that the power curve is all over the place  

[Mono]

I'm not a native english-speaker, but I think it's "rider" (you ride a horse, bicycle, motorcycle, you drive a car). And AFAIK, the balancing of Segway & EUCs are based on the inverted pendulum -equations (see for example here:  http://forum.electricunicycle.org/topic/442-diy-saddled-euc-and-segway-clone-source-code/#comment-4524 )

You can see in @Jason McNeil's Kingsong tests that the power usage is very spiky (this is going uphill at full speed):

I have no other explanation for the spikes except the wheel changing the rotational speed to stay balanced. While the changes may not be that big, they must happen really fast (acceleration/strong deceleration for a split second), so they might require more power than we think.

 

I have a few comments: 

• spiky peaks in itself are not costly. AFAICS, the energy cost is the integral/area and the area in the peaks is rather small, compared to the overall area. Whether a change needs to be fast shouldn't make any difference in the energy balance apart from making the area a little wider in the beginning. 
• any (little) bump in the road should create a peak to keep the speed constant
• a spike doesn't need to be related to change in rotation speed, it's firstly related to change in torque
• some of the observed ruggedness could be artifacts from measuring or a natural occurrence from the motor, independent of the demand
• I can't conclude from the data how energy-costly the balancing part actually is (in particular as I still don't see how to differentiate between balancing and keeping velocity constant)

As non-native speaker my feeling was that a rider should have a seat to ride on 

[Mono]

It matters from a weight and air travel point of view. Today I went to the farmers market. I could have put a 6-7Kg electric skateboard on my back while shopping, and safely made it back and forth. I could not lug a 13-15Kg EUC in the same conditions.

You have definitely exaggerated the weight of a pure e-wheel for short distances without any extra gadgets, which is rather 10-11kg (e.g. GotWay, SoloWheel, TG,...). The current generation of e-wheels hasn't had the focus on saving weight, but on having a cool design and speakers, bluetooth, light, USB port etc. I am pretty sure there will be future generations with much more focus on reducing weight as well.

Otherwise I am entirely with you: for maybe even the majority of use cases of e-wheels in daily transportation, weight is a critical factor.

[robca]

I don't have exact numbers for electric skateboards, but even relatively low end units like the Yuneeq e-Go can do ~25 km range with a 240 Wh battery (and weighs only 6.3 Kg). Can also do 20 km/h, albeit not at max range. I don't understand how people can claim that self balancing won't use more power compared to coasting; it would be like arguing that covering the same distance in a Tesla at constant speed would use the same energy as accelerating and (regenerative) braking all the time. It doesn't. At best, regenerative braking is roughly 30% efficient only (due to the losses in the mosfets, heat, plus the losses in the battery charging efficiency). Any motor when accelerating wastes power in heat (both in the mosfets and the motor windings themselves). Regenerative braking wastes more energy (when charging a battery, only a portion of the energy provided is stored, the rest is wasted. And you can't charge a battery as fast as you can discharge it, so the same energy that can be converted from a battery to a motor, cannot be put back in at the same rate, some must be wasted to prevent damage to the battery.

The graphs Jason published are too coarse to see the real work a motor should put out, and measure the overall current at 10 Hz. If you even looked at the code to self balance a Segway-like device or a EUC, you will see how fast the processor changes the input (at a minimum, 100 Hz, ideally more for smoother running devices... there's a reason why everyone is moving towards 32 bit processors for the control boards). Just imagine how spiky Jason graphs would look at 20x the logging rate (right now that device is averaging power over 1/10th of a second, when the wheel changes power probably at least 200 times a second or more). Every one of those changes are not 100% efficient. It all adds up. Also, as frequently discussed here, a EUC needs a much more powerful motor to provide a smooth ride, due to the need of having a lot of reserve power to balance even at max speed (while a skateboard can use the motor flat out, since there's no need for more power once at max speed)

I'm not here on a EUC forum arguing for the superiority of skateboards  they are not superior in absolute terms, and have their downsides... but electric skateboards are, unarguably, more energy efficient. Rider weight also has a smaller effect on the ride (no need to balance a heavier inverted pendulum). Which means lighter devices everything else being the same. There's another advantage for skateboards: with a dead battery, an electric skateboard (esp the new ones with in-hub motors, which is where everyone will be soon) with a dead battery is just like a slightly heavier skateboard, and perfectly usable to ride on flat and downhill terrain. A dead battery in an EUC means lugging along a heavy device. New designs like the Marbel claim (not independently verified yet) to do 40km/h, range of 25km with a 165 Wh battery... and it weighs 4.5 Kg.

The most interesting thing is that Chinese manufacturers seem to have 0 focus on electric skateboards, so the rate of progress in EUC is currently much faster, and hopefully we will see lighter units soon. In the end, the difference between carrying 5 Kg and 11-12Kg is massive, and that matters to achieve a practical transportation device.

 

 

[Jason McNeil]

@robca, I like your style, well reasoned & presented arguments  

Even with 64cells, the battery pack is only 3kg compared to ~10kg for the direct-drive hub motor. If there is weight shedding potential, it will almost certainly have to come from there. It will be very interesting to see the tear-down individual component weight of the IPS zero, since they claim to be able to produce 800W of continuous power with 32 cells for a total system weight of only 9.3kg, which if true, is nothing short of miraculous.

Perhaps eWheel development will take two different courses: one catering for power-users & one for those who place weight saving at a premium. I could be wrong, but from all the technical materials I've read, I think the two requirements are mutually exclusive, either you have a beast of Wheel or you have a light one. 

The criticism about the capture rate is valid: the hardware allows one to set up to 50Hz, will see if this makes a difference in resolution for capturing swings in power output. 

 

[Mono]

I don't have exact numbers for electric skateboards, but even relatively low end units like the Yuneeq e-Go can do ~25 km range with a 240 Wh battery (and weighs only 6.3 Kg). Can also do 20 km/h, albeit not at max range. 

Range and speed then don't sound much different from EUCs (EDIT: EUCs with the same batter capacity). 

I don't understand how people can claim that self balancing won't use more power compared to coasting;

I don't think anyone was claiming this. The question is: how much of the energy is purely "wasted" on balancing, and, how much energy needs to be necessarily wasted on balancing with the best-case controller implementation, and, whether balancing is a relevant factor in (a) the energy balance and (b) the necessary weight to carry. 

it would be like arguing that covering the same distance in a Tesla at constant speed would use the same energy as accelerating and (regenerative) braking all the time.

However, for some reason we don't notice all this accelerating and braking with our senses. Which seems to suggest that not so much movement is induced in the end, which suggests that not much energy is (at least in principle) needed in the end. (I do understand that the controller operates on time intervals that are not notable to our senses). We know that pole balancing can be done very smoothly with comparatively low acceleration forces applied, if the controller is smart enough and starts in the balanced position. 

The graphs Jason published are too coarse to see the real work a motor should put out, and measure the overall current at 10 Hz. If you even looked at the code to self balance a Segway-like device or a EUC, you will see how fast the processor changes the input (at a minimum, 100 Hz, ideally more for smoother running devices... there's a reason why everyone is moving towards 32 bit processors for the control boards). Just imagine how spiky Jason graphs would look at 20x the logging rate (right now that device is averaging power over 1/10th of a second, when the wheel changes power probably at least 200 times a second or more).

Again, consumed energy is the integral, AKA average, am I wrong? From the graphs I can conclude that going uphill takes about 800 Watt and balancing takes less than 150 Watt (could be 100, could be 10, could be 1), assuming the wheel was balanced between time stamps 0.2 and 0.3. I can't see how to put a lower bound on the "balancing energy" from this data though. I would be delighted to learn how. 

[...] There's another advantage for skateboards: with a dead battery, an electric skateboard (esp the new ones with in-hub motors, which is where everyone will be soon) with a dead battery is just like a slightly heavier skateboard, and perfectly usable to ride on flat and downhill terrain. A dead battery in an EUC means lugging along a heavy device.

That is definitely so and holds true for pedelecs as well. It means that battery size of comparable e-wheels must be larger than for pedelecs or e-skateboards, giving an e-wheel roughly a 1/2kg disadvantage (thereby added 10km safety margin range).

 

The most interesting thing is that Chinese manufacturers seem to have 0 focus on electric skateboards,

For some reason I do understand why  

I never thought for a second that an e-skateboard would be my future transportation aid. But I immediately got suspicious when I first learned about e-wheels. One reason: any device with stand-alone wheel(s) that are smaller than, say, 12" has in my world only toy status, because it can be safely driven at moderate speed only on plain flat surfaces. This is a deal breaker for non-leisure usage, and a bummer for (my) leisure usage as going off-road is the real fun. Another reason: foot placement, which is somewhat unnatural to begin with (fine in sports, less so for day-to-day usage).

 

@robca, I like your style, well reasoned & presented arguments  

Even with 64cells, the battery pack is only 3kg compared to ~10kg for the direct-drive hub motor.

10kg can't be accurate, as a number of direct-drive hub wheels already now weigh less than 11kg, including battery and tire. 

If there is weight shedding potential, it will almost certainly have to come from there. It will be very interesting to see the tear-down individual component weight of the IPS zero, since they claim to be able to produce 800W of continuous power with 32 cells for a total system weight of only 9.3kg, which if true, is nothing short of miraculous.

While I don't believe in miracles, I am not at all surprised by this number. 

[MarkoMarjamaa]

You have definitely exaggerated the weight of a pure e-wheel for short distances without any extra gadgets, which is rather 10-11kg (e.g. GotWay, SoloWheel, TG,...). The current generation of e-wheels hasn't had the focus on saving weight, but on having a cool design and speakers, bluetooth, light, USB port etc. I am pretty sure there will be future generations with much more focus on reducing weight as well.

Otherwise I am entirely with you: for maybe even the majority of use cases of e-wheels in daily transportation, weight is a critical factor.

I'm rather skeptical about advances in weight reduction. The motor weights about 9kg of all wheels. But if they succeed to make that to 5kg it will greatly help mass adoption especially for women.

[Mono]

I'm rather skeptical about advances in weight reduction. The motor weights about 9kg of all wheels.

What's your source for this claim? Does this include bearings and rim?

[MarkoMarjamaa]

Source?

http://www.wheelgo.com/component-weights/

[Mono]

http://www.wheelgo.com/component-weights/

OK, the same IPS now claims to build a complete 800 Watt wheel with 9.3kg weight. You also noticed that it says wheel/motor, which I guess means motor+wheel? It probably even includes tire and pedals and pedal mounts (as seen in the picture), as these are not listed otherwise.