Start with the sensors, then design the rest: How Zoox built its robotaxi

The bidirectional design has some clear advantages for a working taxi.

See full article...

 

[ardentsonata]

This makes me miss the old Waymo Firefly cars even if they didn't go quite as far as what Zoox has done.

Personally agree that you should design the transportation to fit its purpose. If the future is no human driver... Then we should design the vehicles with that in mind.

 

[Cthel]

Does the vehicle interior indicate which direction it's going to set off in when the passengers board? I know some people have issues with travelling facing backwards, which could easily happen by accident if they see the cab drive up and assume it will keep moving in the same direction.

 

[YetAnotherBoris]

Those flat fronts may make a lot of sense from an engineering perspective but they can be extremely bad from a pedestrian safety perspective.

Tell that to pretty much every bus and trolley in the world. (Though for some reason, the majority of American school buses - and possibly prison busses - seem to be the exception.)

 

[Disappointed_In_Ars]

Those flat fronts may make a lot of sense from an engineering perspective but they can be extremely bad from a pedestrian safety perspective.

elaborate?

 

[metope]

I prefer this WestWorld robotaxi design to just taking the driver out of a regular car and call it a Robotaxi.

 

[Jon Tseng]

I haven’t had a chance to try out the rider experience yet, but I’m curious to see how it compares to the black cabs I grew up with in London.

Oh! All these years and I never realised you were a Brit!

 

[sorten]

As any Uber or Lyft driver will tell you, it's also beneficial to have a car interior you can hose down -- particularly in Las Vegas.

 

[eduardopozo56]

a traditional long-nosed hood has a smaller impact area and will send you into the wind shield (where you can be seen). a big boxy front is just a big wall and the driver might not even see you.

https://lloydalter.substack.com/p/study-higher-hoods-higher-speeds

No driver here, so visibility is not a problem. Also, the study seemed to count both speed AND height together, but do not mention (or i missed) height-only changes, it assumed that higher height==less visibility and therefore higher risk.

As others noted already, basically every bus is the world is already flat in the front.

 

[bone_collector]

a traditional long-nosed hood has a smaller impact area and will send you into the wind shield (where you can be seen). a big boxy front is just a big wall and the driver might not even see you.

1. Larger impact area = less pressure applied to any particular part of the body.
2. No driver.
3. Per the article, the sensors are

each perched on a little ledge projecting from the top four corners of the robotaxi’s body. From up there, each has an unobstructed, high-level view, giving the Zoox robotaxi good situational awareness

 

[SirOmega]

I live in Vegas, so I can tell you I'm not a fan of the Zoox. I've seen plenty of videos, and personally experienced twice when the vehicles did something unexpected and/or dangerous (as a driver in a different car, not as someone in the Zoox).

These companies are on the bleeding edge. I wouldn't be pushing expansion if I were them.

 

[Zoc]

I'm a bit worried that the Janus design and odd turning capabilities would make its movements hard to predict for the human drivers it's sharing the road with.

 

[Tactical Finesse]

No driver here, so visibility is not a problem. Also, the study seemed to count both speed AND height together, but do not mention (or i missed) height-only changes, it assumed that higher height==less visibility and therefore higher risk.

As others noted already, basically every bus is the world is already flat in the front.

He's right in a way. Front-over collisions are extremely bad for anyone outside a car. However, almost all of them are Americans driving over-large SUVs and pickups who are careless. Not busses or mass-transit vehicles. Which, in general, isn't that applicable to traditional transit vehicles, like you state. There's an entire class of pedestrian-collisions now, where parents are running over their own kids in their own driveways because they can't see in the near-field.

That being said...robotaxis are making the news more and more as they become more common for misbehaving...because they don't understand what, say, a police barricade is and drive through them. Of course, driverless transit vehicles have been a thing in other countries for decades--they just run on steel-rails with highly predictable pathing and scheduling--and so are proven safe technologies. Whereas Americans and their tech-bros think they can just remove the human from operating a car and it'll get the same end-state.

 

[julesverne]

Sideways facing benches with three point restraints conceived for frontal collisions. Slaps forehead. These vehicles have a top speed of 120kmh. Collision avoidance may have been a top priority. Passenger safety clearly less so.

Addendum: Sideways facing seats in small vehicles are banned in the EU, for exactly the reasons I alluded to. Good safety solutions remained unsolved.

Edit: thanks to @Tactical Finesse and others for the correction. The seats are not side facing.

 

[J Carr]

Each Zoox should have a Gray in it serving as concierge.

 

[Cthel]

Sideways facing benches with three point restraints conceived for frontal collisions. Slaps forehead. These vehicles have a top speed of 120kmh. Collision avoidance may have been a top priority. Passenger safety clearly less so.

Addendum: Sideways facing seats in small vehicles are banned in the EU, for exactly the reasons I alluded to. Good safety solutions remained unsolved.

The seating is transverse (facing forward/backwards), not longitudinal - the doors are on the sides, so that interior render is looking from one side of the vehicle to the other.

 

[jack1983]

Sideways facing benches with three point restraints conceived for frontal collisions. Slaps forehead. These vehicles have a top speed of 120kmh. Collision avoidance may have been a top priority. Passenger safety clearly less so.

Addendum: Sideways facing seats in small vehicles are banned in the EU, for exactly the reasons I alluded to. Good safety solutions remained unsolved.

That was my first thought as well, but I think those are actually front and rearward facing. The doors are on the sides, with the seats at each end. It does look confusing from that photo though. Rearward facing seats are generally safer (almost all baby seats in the EU are rear facing now), so this should be fine.

 

[solomonrex]

He's right in a way. Front-over collisions are extremely bad for anyone outside a car. However, almost all of them are Americans driving over-large SUVs and pickups who are careless. Not busses or mass-transit vehicles. Which, in general, isn't that applicable to traditional transit vehicles, like you state. There's an entire class of pedestrian-collisions now, where parents are running over their own kids in their own driveways because they can't see in the near-field.

That being said...robotaxis are making the news more and more as they become more common for misbehaving...because they don't understand what, say, a police barricade is and drive through them. Of course, driverless transit vehicles have been a thing in other countries for decades--they just run on steel-rails with highly predictable pathing and scheduling--and so are proven safe technologies. Whereas Americans and their tech-bros think they can just remove the human from operating a car and it'll get the same end-state.

Let's celebrate a smaller car by design, then.

 

[aleph_nought]

The people designing and deploying these are tech bros who, somehow, think there's mountains of money to be made in not-mass-transit. Which it shows.

They don't consult proven technologies like trams running on rails that have been driverless for years--that can move a lot of people, very safely, quite efficiently given good transit-law (like transit vehicles should always have superior right-of-way like ambulances and LEOs and firefighters)...instead they think The Future are these small ugly not-quite cars that are designed to look good as a tech demo or from a crappy B sci fi flick.

Dissenting opinion here. I've taken mass transit across the world and I've seen some of the best tram systems in Europe but they don't get you everywhere. There are destinations that still require walking long distances or a short rideshare fare even after a long tram or subway journey.

With a robotaxi, your little party can have a long conversation from the office or your doorstep all the way to the destination in privacy and comfort. Robotaxis won't replace other forms of transit that can carry more passengers but that's not the point - sometimes a hub and spoke model works best, sometimes you need a point-to-point option, and a smart transit system combines both.

 

[MST2.021K]

Oh! All these years and I never realised you were a Brit!

I definitely want a Brit's opinion on this taxi when they get the opportunity, not for passenger comfort but luggage capacity. The cab in London was astounding for fitting my travel group and all our bags.

 

[Tactical Finesse]

Dissenting opinion here. I've taken mass transit across the world and I've seen some of the best tram systems in Europe but they don't get you everywhere. There are destinations that still require walking long distances or a short rideshare fare even after a long tram or subway journey.

With a robotaxi, your little party can have a long conversation from the office or your doorstep all the way to the destination in privacy and comfort. Robotaxis won't replace other forms of transit that can carry more passengers but that's not the point - sometimes a hub and spoke model works best, sometimes you need a point-to-point option, and a smart transit system combines both.

Well you're technically wrong and technically right here. Most of the geographic USA, functionally, has no transit to replace. Decades of austerity have defunded them to the point of practical non-existence, where quality is so bad only those unable to afford or legally operate a car use it. Robotaxis in the USA are being designed and deployed as The Transit option, because outside of the largest metro areas (e.g. NYC, LA, DC, Boston, Seattle) your transit options are generally 3rd-world quality at best.

My Fair City of 300,000 has bus service...but only 6 days a week for less than half the day, with timings of between 30 minutes to 3 hours between runs. Even during the workday it would take over an hour to use the underfunded city bus to go 5 miles--a trip by bike that is 20 minutes. The "bus stops" are literally signs pounded into the easement without a bench or rain cover. That is how bad transit even "good" transit is in the USA where it exists.

 

[Statistical]

The people designing and deploying these are tech bros who, somehow, think there's mountains of money to be made in not-mass-transit. Which it shows.

They don't consult proven technologies like trams running on rails that have been driverless for years--that can move a lot of people, very safely, quite efficiently given good transit-law (like transit vehicles should always have superior right-of-way like ambulances and LEOs and firefighters)...instead they think The Future are these small ugly not-quite cars that are designed to look good as a tech demo or from a crappy B sci fi flick.

Even cities with good mass transit have cars. Even people in those cities which don't personally own a car use cars in the form of taxi and rideshares.

Your utopian world of no cars is never going to happen. Even if there was the political will and money it would take decades.

So if cities will have cars the question is what type and how many. Most of them being BEV robotaxi seems a solid outcome among the real world possibilities.

 

[fellow human]

1. Larger impact area = less pressure applied to any particular part of the body.

Intuitively I would think the legs being swept first would cause there to be a bit less pressure on the upper parts of the body, such as the head, due to a more gradual acceleration. Hitting a vertical wall seems like the worst possible case to me.

Just guessing though, it would be interesting to see some data.

 

[fenris_uy]

because we are really focused primarily on dense urban areas at the moment

Looking at the car, it shows that you designed for dense urban, because that car would not offer that much protection for it's passengers in a highway crash.

 

[tRexx]

I think Zoox has come up with the best design given the mission.

 

[Statistical]

Nice design but the whole self-driving taxi thing just doesn't seem to make any economic sense. Even if the self-driving is perfectly safe (which, sharing the road with a bunch of unpredictable humans, seems unlikely) how is this ever going to compete with a Lyft, Uber, or similar? The cost of labor in a taxi service is just not that high. Plus you're still going to have to employ people to monitor the robo-fleet and clean out all the crap people leave in the taxi. Are they banking on people paying a high premium to take a ride without even the minimal human interaction involved with a app-hailed taxi? As a design/technology experiment it's kind of cool, but as a business proposition it seems a bit daft.

Labor is a significant portion of the cost of an uber. It isn't the only cost but it is significant. Also since uber drivers don't share one car the amount of utilization hours on the hardware is pretty low. Even a full time uber driver might work 50 hours. The other 100+ hours the vehicle is sitting in the driveway making zero revenue but racking up interest and insurance costs.

Fuel is also a significant cost and so on a per mile basis a BEV would be ideal but BEV have higher upfront costs so getting more utilization per day improves economics. The closest thing to the efficiency of a robotaxi would be two guys who split the cost of a mid sized BEV to each do 8 hour shifts using it. That improves economics but I assume that rarely happens. I would add that using a car intended for personal use as a taxi is non-ideal. It has creature comfort and features which added to the purchase price but which produce zero revenue.

Now personally I am not sure the tech is there and it certainly may not be for 5-20 years however once it is robotaxi will absolutely make economic sense.

 

[Tactical Finesse]

Even cities with good mass transit have cars. Even people in those cities which don't personally own a car use cars in the form of taxi and rideshares.

Your utopian world of no cars is never going to happen. Even if there was the political will and money it would take decades.

So if cities will have cars the question is what type and how many. Most of them being BEV robotaxi seems a solid outcome among the real world possibilities.

Tell me where I ever said that was my utopian world? Thanks in advance.

Of course cities will still have cars. But fewer cars is a better outcome--for everyone, both drivers of cars and pedestrians.

 

[fenris_uy]

elaborate?

You get hit at chest level, instead of being hit at the legs or hip. And you fall under the car instead that over the hood.

That's comparing with a sedan. SVU are known to be pedestrian unfriendly because of how high the front is.

 

[Cthel]

Intuitively I would think the legs being swept first would cause there to be a bit less pressure on the upper parts of the body, such as the head, due to a more gradual acceleration. Hitting a vertical wall seems like the worst possible case to me.

Just guessing though, it would be interesting to see some data.

You get hit at chest level, instead of being hit at the legs or hip. And you fall under the car instead that over the hood.

It's not actually vertical though - it looks to be about 10 degrees from vertical, and the front-most point is a lot closer to the ground than the grill of a pickup, which means the first point of contact is going to be between ankle and knee, rather than between hip and clavicle.

 

[fenris_uy]

No driver here, so visibility is not a problem. Also, the study seemed to count both speed AND height together, but do not mention (or i missed) height-only changes, it assumed that higher height==less visibility and therefore higher risk.

There are charts in the article comparing cars vs speed. Bigger cars injure you more than smaller cars at similar speeds.

As others noted already, basically every bus is the world is already flat in the front.

EDIT:
Pedestrians getting hit by buses have a very bad time. The bus design is a compromise into giving visibility to the driver in a dense urban area when maneuvering a big vehicle vs using the less space posible to take the most people vs pedestrian safety.

 

[Statistical]

Of course cities will still have cars. But fewer cars is a better outcome--for everyone, both drivers of cars and pedestrians.

Then robot taxi are part of that future.

 

[fenris_uy]

Of course, driverless transit vehicles have been a thing in other countries for decades--they just run on steel-rails with highly predictable pathing and scheduling--and so are proven safe technologies. Whereas Americans and their tech-bros think they can just remove the human from operating a car and it'll get the same end-state.

Cities with extensive subway, trams and buses lines, still have taxi services on them.