How Lexus Fixed Tesla's Bad Idea: Steer-By-Wire Yoke

Lexus Is Going To Change Steering Forever With Their New Yoke!

The Lexus RZ 450e has a surprising amount in common with the Tesla Model S. They're both electric, they're both all-wheel-drive, and they both offer a steering yoke! Not surprising, Lexus doesn't want to call it a yoke, because Tesla's steering yoke got so much negative feedback upon it's release, and with good reason. Lexus has overcome the Tesla's flaws, however, with a clever new design: steer-by-wire. You see, there isn't anything physically connecting the steering wheel to the steering rack; all communication is electronic!

Before you freak out, know that we'll deep dive into this technology in this video, answering the following five questions:
1) How does steer-by-wire work?
2) Why would you use steer-by-wire, and what are the advantages?
3) Can you trust steer-by-wire to be safe and reliable?
4) Why is Tesla's steering yoke a bad design?
5) What drawback's does Lexus steer-by-wire system have?

Enjoy the video!

Engineering Explained is a participant in the Amazon Influencer Program.

Don't forget to check out my other pages below!

[kallumblake]

I feel one of the biggest benefits is being able to build cars that are either left or right hand drive. As the steering column is one of the hardest things to shift from one side to the other but not in a drive by wire system.

[UnbarablePain]

One important benefit you didn't touch on Jason is that removing the steering column leads to increased crash safety. Steering columns are a real pain for automotive engineers as the linkages needs to be able to break so that a column doesn't impale your chest. Removing the steering column completely helps with crash safety a lot.

[danielrose]

Today there are already plenty of construction and mining vehicles which use steer-by-wire with some great features. One I really like is the angle-variable steering sensitivity instead of a speed-variable one. In the center you get a really precise steering with a high ratio, while towards the limits you get a much lower steering ratio. The same trick is used by several pc racing wheels.

[waxilliumladrian]

Steering Ration differences were already a thing 15 years ago in my car and is a thing in most german cars. It was also adjusting to the ratio to the speed for comfort and those cars have and had a steering shaft.

[reflex]

Not wild about this but I’d love to try one in the real world. I feel like the variable steering ratio would take some time getting used to and could possibly cause some issues when first getting used to it. The ratio could be different than the driver is expecting in some situations when learning the system

[diamondflaw]

My primary concern is how easy they have made it to access and maintain with that many components. In my experience, when things like this get changed, self maintenance is what suffers because it takes a lot of extra work to engineer things to be also easy to replace/repair.

[mikeyc]

I'm glad you addressed the lag toward the end of the video. Within 5 seconds of starting the video I literally said out loud "Wow. You can SEE the lag!" That is a deal breaker.

[linkin]

Your steer by wire subscription has expired, please purchase to continue steering.

[theheadone]

That steering wheel delay alone would be a deal breaker for me. But I do like the idea of adaptive steering wheel ratio (done well).

[declankillam]

The steering ratio that automatically adjusts during acceleration seems really interesting. One thing I would like to try, is to go in a very large open space, turn the yoke a certain amount and then accelerate while keeping the yoke in the same position. It would be interesting to watch the line start off turning sharply, and then it kind of smooths out and gets straighter the faster I'm going.

[jsvn]

for a lot of fps gamers, they turn off mouse acceleration (so mouse & cursor movement ratio is constant regardless of speed) for better muscle memory and accurate aiming. I feel like the dynamic steering ratio will behave similarly to mouse acceleration for cars, it will be convenient to a lot of drivers, but it will take time to get use to, and some people will absolutely despise it.

[AlecThilenius]

A software bug with your accelerator isn't likely to be fatal. A software bug with your steering wheel very likely is. Even the F-22 with an "unlimited" R&D budget lost an aircraft to a FBW software bug. It's just hard to write software that's as reliable as a chunk of metal...

[EngineeringExplained]

**Important Update** Many have asked if the maximum angle the front wheels can turn is limited at higher speeds, since it uses a shorter ratio. I reached out to Lexus for more information. They said "the amount or angle the wheels of the vehicle can turn left-to-right in a vehicle equipped with Steer by Wire is the same at 5mph and 85mph (they are also the exact same angle values left-to-right as on the traditional EPS steering system vehicles). The changes are in the actuation of that steering rack left-to-right and the variable speed that you reach during that turning function as vehicle speed increases."

I have asked some clarifying questions and will provide updates here once received.

**Edit 1:** Here is the clarification they provided. FYI - it does not match what was stated above.
"RZ with Traditional Steering completes a full lock to the L or R turn in roughly 540°.

RZ with Steer By Wire (version in France) completes a full lock to the L or R turn in roughly 150°. This implies the base ratio is 3.6:1, which that for every 1° of driver input at the steering wheel in a Steer-By-Wire RZ, the wheels on the road would operate equivalent to the driver providing a 3.6° input in a traditional steering (540° system).

This solves the question of low speed ratio, but as we accelerate we reach a point closer to 1:1 or beyond, in this scenario if the steer-by-wire driver only has 150 partitions or input angles that can be sent to the computer, at 1:1 ratio – the Steer By Wire system only has 27% potential of available max steering angle (equivalent to roughly 1/3 of a full-lock turn).

HYPOTHETICALLY SPEAKING FROM HERE ON: As we approach exceptionally high speeds, the ratio could become even greater skewed on Input:Output so 1 degree input could mean 0.2 degree of output for the wheels. That means that for every 5 degrees of steering wheel turn by the driver, the wheels at the road would be turning the equivalent of 1 single degree input in a traditional steering vehicle. This provides the feeling of ultimate precision and fun while driving at high speed because you are truly feeling every decimal partition of steering angle so you can very accurately steer the vehicle where you want it go.

540° & 150° are the angles we know and have experienced. The 3.6:1 ratio is a simple calculation that shows the relationship of those values. We have a confirmation but not a provided data sheet, that at some speed beyond 10mph the ratio begins to skew towards 1:1 or more, so the rest of the values (i.e. 0.2°) are completely made up values but the math proves the concept.

Purely for the case of the argument – let’s say the driver is doing 80mph and at that speed every angle of driver steering wheel input is 0.2 equivalent degree of output at the wheels. If we multiply the 150 degrees of full lock input available to the driver at the steering wheel by the 0.2 degree output at the wheels, our total steering angle would be 30 degrees.

As the driver applies the brakes or accelerates, the algorithm of input/output adjusts thus giving the driver “more access” or “less access” to max steering angle.
5mph: 150° Input by 3.6 Output = 540° “full lock turn” equivalent or 100% of max steering angle available
45mph: 150° Input by 1 Output = 150° equivalent or 27% of max steering angle available
60mph: 150° Input by 0.35 Output = 52.5° or 9.7% of max steering angle available
80mph 150° Input by 0.2 Output = 30° or 5.5% of max steering angle available

This may seem strange to think about, but imagine going 80 mph and without slowing down turn the steering wheel of ANY vehicle 540° to the full lock position…… nobody in their right mind would ever do that because it would undoubtedly cause a serious accident. Having access to truly full lock steering ratio at 80mph is not something anyone ever takes advantage of or is physically capable of doing on a road because as you begin to turn that aggressively at speed, you would completely run out of road surface to drive on before being able to even complete the full lock turn. There is also the high likelihood that if a driver attempted to do a full-lock turn at 80mph, the vehicle would have the possibility to flip over. We know it takes about 2.2 seconds for an average driver to make a 540°. At 80mph, approximately 200 feet are covered in 2 seconds. With the average lane width in the US being 12ft (source: Google Search), it would require a road 16.67 lanes wide to be able to complete a full lock turn (without slowing down at 80mph, if that were even physically possible without disrupting or flipping the vehicle over).

Conclusion: The higher speed one goes in ANY vehicle, traditional steering or Steer By Wire, the less useful max steering angle becomes, if even physically possible to achieve. So the Steer by Wire system mathematically limiting access to max steering angle becomes a non-issue from a drivability or experience standpoint. Having an inherent base ratio of roughly 3.6:1 and (a hypothetical) 0.2:1 at high speeds provides extreme ease of use at low speeds in parking lots or u-turns while providing an exceptionally precise feeling when steering at higher speeds and provides a complete range of differing input:output ratios to match the situationally indefinitely. (The faster you drive, the more precise the steering ratio)."

[arxaaron]

When left as an OPTION, this seems really smart. Based on your comprehensive evaluations, I expect I would be very uncomfortable with steer by wire, so would have problems with being forced into using it.

[cabooseledgend]

My concern with the speed relative steering sensitivity is how it would behave when you slow down quickly.

Example being if you're going along a bend on the highway and suddenly traffic stops in front of you or something.

If your panic response is to slam the break and grasp the steering wheel at the angle you were holding it, then as the car slows it would steer harder and harder into the direction you were turning.

That might be something we would intuitively compensate for, but I feel like it could also throw some people off and cause an overcorrection or loss of control.

I'd love to see someone test a scenario like this and report how it behaves.

[ei..]

The part that excites me the most is the adjustability. I'm "only" 6'3", but I've yet to drive a car where I can be far enough away from the pedals while still reaching the steering wheel comfortably.

In my current car i have the seat as far back as I can, but with the back very straight to push my upper body closer to the wheel.

[telpeloth]

FYI steer-by-wire was already present in 2013 Infinity Q50 series.
Yes, ten years ago. It had a fallback to a direct driving shaft via a clutch

[ivizaxyo]

I would love to try this on the real world. My only "concern" really would be how it performs on the moose test, avoiding obstacles at high speed, and how would it deal when you approach a roundabout while going quick. Aside of that the tech sounds solid.

[albertojoseyanespantin]

Fly-By-Wire has been a thing in aviation for quite a while now, it's interesting to see it now introduced to cars.

[Tarets]

We've been dealing with these problems for years in simracing gear and although they do feel great now, they still don't feel like the real thing, and that's despite having the feedback calculated almost instantly, rather than measuring the forces with sensors. That is because real wheels introduces constant resistance, whereas electronic ones can provide a single-directional force only, that is applied as a response to driver turning the wheel - you need to turn first before the motor kicks in, which makes it feel disjointed.

It's also terrifying to see the wheels decalibrate at 12:35, especially considering all these multi-redundant safety claims.