What is the best for robotics: Electric vs Hydraulic actuators

To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/Skyentific/. The first 200 of you will get 20% off Brilliant’s annual premium subscription.
I have looked at the hydraulic actuators (from the theoretical point). I received a lot of comments in the past that I should try hydraulics. I know nothing about it, so I decided to start by comparing hydraulics with electrical actuators. And both of them have their pros and cons.
This video was sponsored by Brilliant
Please subscribe. This will help me to develop other projects like this, to bring the bright future closer!
One time donation:
www.paypal.me/Skyentific
If you want to help this channel, please support me on Patreon:
/ skyentific
Instagram: / skyentificinsta
Facebook: / skyentificface
Twitter: / skyentifictweet
#robotics

[@Skyentific]

To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/Skyentific/. The first 200 of you will get 20% off Brilliant’s annual premium subscription.

[@fabiocroldan]

Can you make a video that compares two electric actuators against a hydraulic one?

[@gatoninja4387]

IT ALWAYS FALLS INTO THE SAME MISTAKES COMPARE PEARS WITH APPLES THE HYDRAULIC SYSTEM IS A JAR WITH OIL LIKE A SYRINGE..AND IT IS POWERED BY A GASOLINE OR ELECTRIC MOTOR AND YOU SHOULD COMPARE THIS JAR WITH OIL WITH THE GEAR REDUCTION SYSTEM OR A CYCLOIDAL OR A PLANETARY ONE OR A PULLEY ONE....PLEASE COMPARE. REDUCTION SYSTEMS WITH REDUCTION SYSTEMS THE STUDY CARRIED OUT BY PSEUDO-SCIENTISTS IS CONFUSED AND POORLY DONE AND LACKS OF INTELLIGENCE. SUMMARY... AN ELECTRIC SERVO CAN BE PUT A HYDRAULIC REDUCER OR A GEAR REDUCER OR A PULLEY REDUCER.. . A GASOLINE ENGINE YOU CAN PUT A HYDRAULIC REDUCER OR A GEAR REDUCER OR A PULLEY REDUCER.. THE STUDY THAT EXPOSES MIXES EVERYTHING AND GIVES A WINNER AND THIS IS A LIE IT IS FALSE TO COMPARE YOU MUST PUT THE SAME MOTOR WITH THE DIFFERENT TYPES OF GEARBOXES WHATEVER BE...AND THE WINNER IS THE ONE WITH THE FEWEST NUMBER OF PARTS THAT CONSUME ENERGY WITH FRICTION AND THE FORCE IS ALMOST THE SAME BECAUSE IT IS THE SAME MOTOR THAT GIVES ENERGY TO THE SYSTEM REMEMBER THE HYDRAULIC SYSTEM IS ONLY A SPEED REDUCER OF THE GASOLINE OR ELECTRIC ENGINE THE SAME AS THE GEAR REDUCER ONLY THAT IT IS HYDRAULIC.. THEREFORE THE GRAPHICS AND THE STUDIO ARE BAD ------------------------LOOK AT THIS EXAMPLE THE HYDRAULIC ARM LIFTS 10 KILOS BUT IF WE PUT A PULLEY CONNECTED TO THE SAME ARM AND WITH THE SAME LEVER TO THE SAME MOTOR OF THE RC EXCAVATOR, IT WILL ALSO LIFT 10 KILOS AND IF WE PUT A CYCLOIDAL REDUCER OR TO THE SAME MOTOR OF THE RC EXCAVATOR OF COMMON GEARS WILL LIFT THE SAME 10 KILOS IN THE SAME TIME... JUST CHANGE THE ENERGY LOSSES FOR THE CHOSEN SYSTEM, IT WILL NOT LIFT MORE BECAUSE IT IS THE SAME MOTOR WITH THE SAME ENERGY kzbin.infosIKYO3WKF6k

[@kingofcastlechaos]

I can walk into a R&D lab and tell you in seconds they use hydraulics by the smell and feel in the air. No matter how careful you are, your lab space gets covered in fluid. One tiny pinhole can spray fluid over the entire lab. Thankfully, books, ledgers, notepads, work orders, paychecks, and other paper items in the space will help soak it all up.

[@Skyentific]

Good point!!!

[@PiefacePete46]

I worked in an airline operations office for many years... You could tell if an engineer had been working on a hydraulic system within a couple of seconds of them coming through the door! That is a unique aroma!

[@samanthaqiu3416]

Do not forget the very impressive robot hand designed by the polish CloneRobotics, it is actuated with air pneumatics and relatively low development budget

[@Skyentific]

I agree that it is extremely impressive!!!

[@samanthaqiu3416]

@@Skyentific I think it is actually water, not air, my bad

[@gordonpromish9218]

@@samanthaqiu3416 iirc, various iterations have used water, a light fluid/oil, and air. I am unsure what their current demonstrator uses - I am not even certain it is a strict hydraulic/pneumatic McKibben actuator at this point.

[@viniciusbembea]

​@@gordonpromish9218 At this point, as 2023, Clone have a hydraulic McKibben actuator for their hand

[@sharg0]

If I'm not mistaken "crank-type radial piston" is basically the same as the radial piston configuration that was used for many aircraft engines (before gas-turbines took over for high performance). A radial engine can be in two basic configurations, rotating pistons or rotating crankshaft.

[@PiefacePete46]

I think a major deciding point is "Why are we making this device"... If it is for an industrial application, the extra cost, complexity, maintenance, and disagreeable 'messy-ness' of hydraulics may be acceptable. If it is for hobby-type use, a totally electric system is usually a much more attractive option, if it can achieve your goals. Thanks for another thought-provoking video.

[@mr.thomas6128]

Also Hydraulic oil is not that easy to get. Electric motors are. Hydraulic cylinders can not be directly printed. Gears for motors can be

[@rtwas]

@@mr.thomas6128 This is incorrect. Hydraulic oil is trivial to obtain. You could use atf (for experimenting) in a pinch but it's pricey and smelly. Standard hydraulic oils are particularly abundant locally (think Tractor Supply, NAPA, industrial oil supply) and can easily be had online. You can 3d print gears with standard 3d printers but high torque, *accurate*, and compact shafting has been an issue, at least for me. I find I have to use square shafts in 3d printing, with vertically printed adapter collars (for accuracy). Everything has to be oversized. Standard 3d printing for rotational machinery is an engineering discipline unto itself IMO. I find I have to work very hard to convert conventional machined transmission parts to the 3d printed world. Hydraulic cylinders on the other hand are trivial to machine with a simple garage lathe. Once can just use pneumatic valves in hydraulic mode in some cases. The hardest aspect to hydraulics (IMO) are the servo valves.

[@savvysymbiont]

This is precisely the research I am looking for, on a new truck-crane concept I am designing. Thank you Skyentific!

[@baraka99]

Robot engineering is my new love. I already have fundamental knowledge of machine learning and python. I'm interested in integrating both in a protoype even if only in a conceptual level.

[@gordonpromish9218]

there are so many factors to consider - output, efficiency, incidental noise/vibration, heat, durability, complexity, robustness, mass, volume, speed, speed variability and control, repeatable precision, compliance, etc... and each class of use will prioritize them differently, with different minimum demands - industrial, commercial, domestic, sport, etc... it is so easy to get lost in the weeds, as a layman.

[@vjorgic]

Hydraulic system can hold the load without the brakes, look into hydraulic helical rotary actuators these are a unique mix of hydraulic motor and cylinder, very useful for robotics application

[@synchro-dentally1965]

I look forward to seeing electric motors that use Iron Nitride magnets. Their magnetic field strength is theoretically more than double that of neodymium magnets. However, the operating temp. range is much lower.

[@patrickderp1044]

i really like this channel and have been following it over the years. if you put out quick videos of printing random knicknacks i would watch that!

[@lucamackenzie516]

Great video, thank you

[@renanmonteirobarbosa8129]

I wish you had properly referenced the papers on the description

[@sambitpradhan93]

Thank you for the great videos sir. You are an inspiration to many. Keep up the great work.

[@hibariba]

How about rolling diaphragm pistons? Disney research has a great demonstration on these cylinders as closed loop direct power transmission

[@davidelang]

a couple other advantagea of hydraulic systems you only need one pump to run many motors due to the higher torque, you may be able to avoid gears. requires no power to hold position (just close the valves, and you only will move to the extent that your system is leaking) high power linear pistons they only make sense for heavier systems with lots of things to drive, which is not the typical hobbiest system,

[@Clark-Mills]

Trivia: SpaceX's starship / super heavy is cutting over to electric from hydraulic for their grid-fin actuators; mentioned something like lighter & less complicated.

[@johnathanclayton2887]

They are also switching for engine gimbals

[@yakut9876]

Electricity is much more complicated than hydraulics.

[@Kodemaestro]

Interesting... I've never looked into hydraulic motors but I've always assumed that they used a rotary vane mechanism, like a rotary vane vacuum pump, but in reverse obviously.

[@DimaKoovikin]

Отличное видео, спасибо!

[@Skyentific]

Спасибо что посмотрели и прокомментировали!

[@kiftesfalem2907]

Hello, what 3D printer did you use to make SmallRobotArm ?

[@linkhidalgogato]

something that u didnt talk about at all but i think is a huge factor is energy storage with a hydraulic system u could have an ice engine driving the pump which is a lot more energy for the weight than most older batteries technologies and since the goal with robotics is often something which is self contained thats very important. im pretty sure lithium ion is still less energy dense but judging by the fact that most newer projects are going electric i guess most people who are actually in the know think the extra energy density is no longer worth the complexity. also i think the percived difference between systems that use hydraulics being more advanced but electic being more common is simply a by product of the fact that hydraulic used to be a lot better so they were used for older systems when they were originally designed and those systems have had more time to develop plus we only consider the success which adds to the whole situation. overall i think hydraulics are just not it anymore, at least when it comes to motors, especially not when it comes to a diy or smaller scale thing.

[@yakut9876]

For me, I prefer Hydraulic and Pneumatic systems. I think there is potential for miniaturization, but companies don't want that ! Because they want to insert electricity into everything, even those that do not need electricity, because it is a consumer system and dependent on them, so through it they control people centrally !!

[@TechnoGlobalist]

I think there will be a split between heavy duty industry with hydraulic (generally bigger robots) and human interface with easy to maintain, weaker humanoids.

[@Skyentific]

This is really great point!

[@ChrissH94]

Nice Video. I‘m researching in the Field of hydraulic Control. Could you gibt me a Link to the Research Papers?

[@yakut9876]

For me, I prefer Hydraulic and Pneumatic systems. I think there is potential for miniaturization, but companies don't want that ! Because they want to insert electricity into everything, even those that do not need electricity, because it is a consumer system and dependent on them, so through it they control people centrally !!

[@heartminer5487]

Crank type radial hydraulic motor is very much like the internal combustion engine design.

[@Datensaku]

You should take a look at the works of Clone robotics, they made a robotic hand that move very similarly to human's hands

[@AGI.ROBOTICS]

Bonjour Tu peux faire un vidéo sur le "GIM8115-36" CODE de programmation et caractéristiques et merci

[@kantorobo7718]

I hope you are doing okey, we havent had new videos hopefully you are in good health :)

[@pecilijevelicanstveni4973]

Depends of the application.

[@phitsf5475]

Amazing coincidence, I'm in the middle of designing a hoist which is actuated by a single hydraulic ram. Operator has two buttons available to them UP and DOWN, pressing either will demand a solenoid to open or close at full speed. What options are available for creating a "ramp" in the output load of the hydraulic cylinder? I can imagine some valve that opens gradually over a period of time but I don't know if that exists in real life. A colleague suggested ramping the pump speed, managing/manipulating motor RPM (and therefore pump output pressure) is apparently trivial.

[@automateTec]

A pressure-compensated adjustable flow control valve can offer smooth ramping manually and could be actuated mechanically. Interestingly the output can be taken from either output ports once the flow is split.

[@H34...]

Proportional valves are what youre after but they are usually quite expensive.

[@RoboSidekick]

what about Dielectric Elastomer Actuators and other attempts to make artificial muscles? on our channel we are building a robotic arm with a HASEL actuator. It is an interesting technology, but it is needs a lot of development and improvement. for now we are thinking to make an end effector based on a HASEL actuator. the advantage here is the actuator can be a tactile sensor at the same time. o

[@zajlord2930]

i have been working with electric motors but i felt like switching to pneumatic/hydraulic might be better not only because of the force, which of you want as much as possible, but also the technology. if you want to make the robotic arm you were working on here on youtube, pneumatic/hydraulic pistons along the arm will have much easier time with it then electric motors in the joints + they wouldn't be part of the skeleton thats holding it together so it would be less likely to fail. but i might be wrong, thats just how i see it edit: i was talking about linear pistons

[@koooooliooooo6399]

I think you are right, that electric motors will surpass the hidraulics in time, making the latter obsolete. In the meantime I guess they could be used together. Lets say an arm. Around the shoulder we need more power, thuss hidraulics, and around the wrist we can use electrical driven system for fast and more precise movements. Also, I noticed that you didnt commented on the wires vs hoses power transportation. I think wires here are better because hoses for bigger pressures tend to be stiffer, thiss limiting movements or forcing us to think of additional movable seald joints, which is more parts -> more problems.

[@yakut9876]

Electricity is much more complex and has many more parts, and electrical wires can break ( Cut ) while Hydraulic components can be made to be very flexible. Otherwise, it is much worse since electricity always needs external power, without a battery, for example, which has a limited lifespan. Often, for every charging and discharging cycle, the system is of no use, but the Hydraulics can control it without external power!!

[@sandorphoenix]

Why not combine them both?

[@buidelrat132]

Advances in modeling, power delivery, and machining should make axial flux electric drives preferable for all but the biggest loads.

[@Skyentific]

I've heard a lot about axial flux electric motors, but I cannot understand why they are not widely used now. May be their are not as good... Because it is relatively easy to make them, I think it is even easier than standard brushless motors...

[@MrTrilbe]

@@Skyentific IIRC Axial flux motors tend to have a larger diameter but shorter depth/height than a more standard brushless and then there's the costs of retooling production lines and redesigning products to use them, each type has it's best use case, axial flux in electric vehicles because the motor can be the diameter of the wheel but is constrained by the width of the wheel due to tire standards for example

[@H34...]

It is because of the moment load they put on bearings as opposed to radial in a normal motor. The forces The stiffness keeping the two axial faces from deflecting and rubbing is much lower. Think you have two parallel plates like in an axial flux motor, you can take your fingers and pinch on the rim and bend the two plates together very easily and make them rub. If you tried to do the same on a radial motor, to pinch the outer shell of the motor to the inner core, its much harder. The magnets and coils are constantly pushing and pulling on motors so you can see how this can deflect an axial motor and cause problems. Bearings also hate moment loads like would be caused by axial plates deflecting. Radial loads are fine though like in normal motor.

[@renanmonteirobarbosa8129]

This is so heated and complex debate that you would need a Phd Dissertation to cover all the details.

[@colinericburriss]

Pneumatic actuator?

[@Skyentific]

As far as I know, pneumatic actuators have one big problem: the gas (air, or any other gas, which is used in such actuator) can be compressed. This is why such actuators will be very springy (the stiffness will be low). And the only way to increase the stiffness is to increase the working pressure. And for the reasonable stiffness the pressure in the system should be really high. I think this is manageable when the you control the piston from one side to another (from one limit position to another). But when you want to put the piston in the intermediate position, it is difficult to get the high stiffness.

[@SimonBauer7]

​@@Skyentific agreed, due to this they are often used in the grippers of industrial robots.

[@ulforcemegamon3094]

Depends of the aplication, if you are building a legged robot electric is the way to go , if you are building a robot arm you can use either Hydraulics or electric

[@BarryGrumbine]

Perhaps it is easier to centralize the weight of a hydraulic system. That would give hydraulic an advantage for humanoid robots. Hydraulic systems, like hybrid vehicles, look as though they are more complex and require more maintenance at a higher price.

[@heartminer5487]

Let's step back a little bit. The real discussion is not with hydraulic or electric motor, but with fluid (piston) or rigid (harmonic/rv) speed reducer. Hydraulic actuator can eliminate pmuch all of the downsides with direct drive valveless design. But the real advantage with hydraulic come from the higher load capacity with given form factor.

[@user-li8zm5kq1j]

А есть русская версия канала?

[@foxtrotunit1269]

Too bad we don't have other options closer to human muscle's power/weight + responsiveness. A *cool concept* would be to have a tiny piezo-electric building block that shrinks by 10% when powered. You chain up 1 000 of these small building blocks and you have an actuator that's presumably both *strong/to weight* and *responsive*

[@soothingunboxing7129]

Mckibben Muscles.

[@foxtrotunit1269]

@@soothingunboxing7129 Ah yes, very cool concept. Problem is it pneumatic. Piezoelectricity (or even standard actuators) only require voltage. Less parts, less problems right? :) (still very cool though)

[@soothingunboxing7129]

@@foxtrotunit1269 also see Fishing Line Muscles. Btw, do you have link to any paper to the piezo muscles?

[@foxtrotunit1269]

@@soothingunboxing7129 It's a concept developed by Dr. C.Halsey of Office of Naval Intelligence. She is a fictional character of the game Halo: Reach. Reason I mentioned is because it's possibly within modern science's reach - and for the fist time we have practical AI for robots, *so for the first time a need for such actuators.*

[@jchestnuts]

Everything that uses oil will become messy and contaminated. Not what I want in a humanoid robot. Electric wins by a mile. Hydraulic is perfect for bulldozers.

[@nou5440]

5:32 that sounds kinda sus

[@SimonBauer7]

i think eventually electric will win, hydraulic is good in industrial applications but everywhere else its terrible.

[@eelcohoogendoorn8044]

Im kinda sad that teslabot did not aspire to make a cheap and affordable hydraulic leg system. Not that I know how to do it; but I do know that electric just does not quite cut it for the requirements of a legged humanoid robot. Every robot built over the decades has shown the same, and the teslabot is no exception; nor will future iterations on the software or hardware side make it much different, barring some fundamental revolution of electromagnetic drive systems, which I do not see on the horizon.

[@ulforcemegamon3094]

The Main problem is always maintenance , Hydraulics aren't that good when powering legs , Atlas needs maintenance as soon as it finishes doing their stunts, BigDog was discarded by the military due to the fact that repairing the complex hydraulic system for the legs was very difficult, the OSU adaptive suspension vehicle had problems with needing constant maintenance due to the hydraulics they use . Due to all of the examples i gave is likely why they went full electric and not hydraulic of any kind

[@Ltanti-furry]

came here after the new atlas

[@MyNutsItch0]

Frrr

[@ibrahimkocaalioglu]

Very informative video. The problem is the experts in these technologies are quite apart from each other. They should learn one anothers technology and come up with hybrid systems for the best performance.

[@maxekb77]

Можно просто учебник почитать, теже тросики не так и плохи. А с нынешними энкодерами так можно сказку творить ;)

[@yakut9876]

Of course, The Hydraulics are BETTER than Consumer electricity and unnecessary complexity.

[@AABB-px8lc]

Responsiveness is upside down obviously, less is better (latency). And one cannot use hydraulic in vacuum (space, DUV lithography ) at all.

[@Jkauppa]

k.i.s.s. dont do what you dont need to be doing over complex

[@Jkauppa]

screw that, servo motors, use diy brushed worm gear linear actuators, they lock

[@Jkauppa]

try electric wire resistance heating hot air blimp drones

[@Jkauppa]

most of your brushless dc motors are brushless ac motors, you got it wrong

[@Skyentific]

They are brushless DC motors :). I never used brushless AC on my channel.

[@Jkauppa]

@@Skyentific remove the control inverter/ac/pulse circuit, watch if they work with dc, definitions lol :)

[@over2seeyer]

as someone studying humanoid robotics, it saddens me to realize that electric motors are no where near muscles in terms of power to weight ratio. secondly, while they they may be lighter then their hydraulic counter part 1 vs 1, its very different when you system has many actuators, you only need one pump and reservoir to drive many actuators/motor while you need just as many motors as you have joints to move, all in all they hydraulic end up lighter

[@christophegroulx7816]

You're wrong, by weight electric actuators have a much higher power/weight ratio

[@MegaTraxxas]

This man has russian accent 😅

[@user-bb9ur9wq3n]

Человек "работает " на гидравлике и малой электрике ) , так что мы уже почти "юсб порты".