Amazon’s electric Rivian vans have delivered 150M packages | Hacker News Reader

Amazon’s electric Rivian vans have delivered 150M packages

arstechnica.com

112 pointswffurr118 comments

I saw one of these a few months ago when visiting Tampa. I was really surprised there was zero Rivian branding on the outside.

More generally, electrifying these sort of vehicle fleets makes a lot of sense since usage is predictable, order volume is high if replacing a whole fleet, they run on a schedule, and the problem set can be well defined (compared to passenger vehicles where you need to solve for every edge case).

I'm a bit more bullish on it. It makes some sense in dense areas, I think. Which almost certainly accounts for the majority of areas that are covered by delivery.

However, the range is not to be discounted. Assuming a delivery van will operate for basically the full working day, I'd assume 200 miles is not unheard of for driving, and I see the range is about 150 miles for them.

That said, https://www.cnbc.com/2023/01/14/what-its-like-to-deliver-for... seems to indicate that the range is working out ok? I'm curious if that is strictly in the dense areas? Mayhap the lower density also translates to fewer deliveries such that you have a lower "duty cycle" as it were?

could potentially supercharge them over lunch breaks if needed. fleets could also battery swap with less complication than normal end user cars.

they also have to be WAY easier to drive. they just scoot out of there rather than having to shift 9 times to get up to speed.

Not just lunch breaks. They have to return to the warehouse to reload every couple hours.

> they also have to be WAY easier to drive. they just scoot out of there rather than having to shift 9 times to get up to speed.

All Amazon delivery vans are automatic.

Agreed, they'd outperform in stop and go driving. Delivery in urban areas is their sweet spot. Since the workload is largely over the work day, supercharging needs are probably limited and vans can be just overnight on level 2 chargers.

Also considering that it's policy UPS (maybe others?) to turn off the engine at every stop, and starting a vehicle is fuel intensive in an ICE vehicle

I don't think that is correct, many vehicles now have the auto start/stop as a way of meeting emissions requirements.

If it's so intensive, why is auto-stop/start such a common feature on many vehicles? Especially since it seems to improve city-cycle mpg ratings...

Auto stop/start requires some (mostly minor) modifications to the engine, as well as updated software so that it is used only in “ideal” conditions. And the engine shutdown is done in a way that makes it quick and easy to restart.

I’m not sure, is it even possible to use these systems on diesel motors?

Of course - I've had a diesel Nissan Qashqai with a start stop system all the way back in 2013. Never had any issues with it.

This is an amusing place where most of the advancement was probably in the starter. The same thing that lets you have a small charger for your laptop has helped get a smaller starter that won't drain after a few starts. There is a lot of work that goes into starting an engine.

It takes less fuel than most people think.

This is almost certainly one of those things where the tech has moved on. It definitely used to use a fair bit of fuel to get started. Not so much anymore.

Probably also depends on where you are? I'd imagine colder climates need to keep the engine running.

Even way back when it used a lot of fuel to get started, it only took a few seconds of engine off the make up that fuel.

I've had cars need jumping from too many starts draining the battery, though. So, mayhap it was less the fuel and more the battery/starter?

Modern cars will disable stop/start if the battery is low. My car won't even use the start/stop feature if the HVAC is operating at a 100% of it's capacity.

> to turn off the engine at every stop, and starting a vehicle is fuel intensive in an ICE vehicle

This is backwards. The whole point of these systems is that it costs more in fuel at idle than it does to stop and restart.

Lots of "pure" ICE cars have this now with some smarts about operating temperature, battery etc.

Idling was worse on carbs, sure, but even with injectors your fundamental problem is you don't have enough airflow, and you counteract that by richening the mix so you don't stall.

Not actually true.

https://en.m.wikipedia.org/wiki/Start-stop_system

Great link and a good resource for folks curious on this. I had posited that it was likely changes in the starter getting more efficient that has caused this to become more common. However, the article makes it clear that it is largely changes in how things are powered in the car. The specific point about the belt being used to power AC and such is likely the largest factor on why this is slowly going out. (And why it is more common in hybrids, since they had to be designed to power all systems without the belt.)

Is a very interesting case study in what software folks would call "tech debt."

> since usage is predictable

I think this is the important bit. Amazon knows on exactly which routes a full battery and a full load of packages line up, so they can put the EVs on those routes.

Another thing I was just thinking of this morning was that you could implement possible charging into a route of packages -- pushing an automatic break in for a driver to stop for 10-15 minutes and grab a DC fast charge somewhere. It'd almost be a win-win: the driver gets a predictable break on the route (and knows they get a break!) and you can keep the vehicle on route longer.

On top of that, while plugged in, they can keep cool/warm in the vehicle without having to idle it or have additional wear and tear...

It ultimately doesn't pan out that well. Will work right up til someone finds that they can get a % more delivered without doing the stops.

You can see this in long race formats that have breaks. Teams that find a way to not do the break wind up winning. Such that you then have to mandate them, which works for things that people don't get value out of.

We already have similar mandates for truck driving, and I think most folks would be surprised that that exists. And they are far looser than what would be required for last mile driving.

> Teams that find a way to not do the break wind up winning

Teams that don't take the break will be inhibited by the limited range of their un-recharged vehicles. Are you suggesting that some people will find out how to swap their vehicles, and keep the shift going? (Or add an extra length of hose to the fuel line, for a virtually bigger tank? That type of trick isn't going to work well with battery vehicles...)

How is it you propose that someone is going to get more work done in less time, when their total range / route length is reduced by the extra effective distance they would have been able to continue with 15 minutes of fast charging added in, at the one or two points where it makes sense to add a stop in their shifts?

Mandated breaks being what they are, you are going to find it hard to codify it in such a way that fits all locales, ultimately. Some routes will just take more power, for reasons.

Then, later, they would find that the newer batteries actually can get by with smaller breaks. Or other efficiency gains that we just aren't thinking about.

My point wasn't that it would fail immediately, but it would eventually.

> How is it you propose that someone is going to get more work done in less time, when their total range / route length is reduced by the extra effective distance they would have been able to continue with 15 minutes of fast charging added in, at the one or two points where it makes sense to add a stop in their shifts?

By throwing the furthest of their deliveries into the nearest lake, and coming back early with an empty truck and charge remaining.

Drivers don't want to sit around waiting for a recharge. They want to clear the truck and go home.

That's not an optimization, lol, and it's not an answer. If the 20 minute recharge was built into a break, with a fast charger, that's going to add an extra 120 miles onto the delivery route range. IDK how many miles Amazon routes are expected to drive in an 8-10 hour shift, but given that the top range of the vehicles is about 350-400 miles with the biggest battery, that's almost an extra 35% of range.

If the trucks go far enough, and the drivers don't need breaks, then say that I guess. Drivers don't want to do their job is not an engineering problem, whereas delivering more packages per shift could be one.

Amazon workers don't get breaks. They work until they break.

Bearish?

Ha, yeah, I used the wrong term. :( Apologies on the confusion.

My main point is still on the "contra to my view" side, though. I'm skeptical that they are as good of an idea as stated, but evidence is positive on them.

You ride a bull, you run from a bear.

Nice! I actually knew the correct term, though. Have a bad habit of switching binary things a lot. :(

Edit: Amusingly, I think I was thinking that I was bullish against it. Such that I flipped a different part of the sentence. Still, wrong as written.

Or: bears hibernate, bulls charge

It's confusing, because people generally don't ride bulls. Where I'm from you gently back off from both bulls and bears.

I remember it as bears hibernate, so they will stockpile resources and not let go of them.

I don't really have an analog for the bulls, oddly.

I was taught it was how they attack.

Bears: stand and swipe down with their paw

Bulls: start with their head low and jut upwards with their horns

DPD does all their deliveries in my area(medium size city in the UK) with Nissan NV100 electric vans and they get 100 miles of range on a good day(more like 60-80). The biggest chunk of that is warehouse to the city, then in-city driving makes very little impact on battery range since you're starting and stopping constantly(which is actually directly opposite to the efficiency of ICE cars)

I'm curious if they have to have more vans than they would otherwise need, then? That could easily let you cover the day with lower range.

I'm also open to challenges in my thoughts on typical range needed for a city fleet in deliveries. Will try and take a dive to find more data there.

I feel like loading whilst charging would minimize the impact charging makes quite significantly. Fast charging takes what, 20 minutes? I'm not sure loading a van is much faster.

Truck operators already do that - Volvo was saying that in the trials for their fully electric truck, charging while the truck is being loaded at the depot is fast enough for all day of operating. Of course that's not long distance driving, but plenty of customers for this kind of thing already.

Another benefit is batteries degrade with age worse than they do with use. So high mileage applications are perfect, way less stuff wears out with mileage than a gasoline/diesel truck. A delivery van can put enough miles on to wear the chassis out before the battery pack degrades too much from age.

I drove pure electric (bmw i3) for 7+ years, and i'm unclear on what "age" means here, since recharge cycles I thought of as defining age, combined with the ambient temps of the battery pack during discharge and recharge.

Rechargeable battery wear can be separated out into (1) cycle aging and (2) calendar aging.

One of those typically dominates the total battery degradation, depending on the use case. EVs for personal use tend to have calendar aging dominate the overall battery degradation early in their life.

The other reply got it right, but I would add that the higher the temperature the faster the pack degrades, even when it is just sitting there.

How do/did you like it? I'm suppose to test drive one next week.

i3 pure electric was great but they dont make them anymore

For the scale they are operating at, it wouldn't be crazy to pull batteries that dip below a set kWh/lb value and use them as electricity storage.

From a business perspective it makes sense if the TCO is lower than an equivalent gas or diesel truck. That's really the only thing a sane business will think about. There's probably some "goodwill" value in the public perception of driving around in electric trucks, but the bank won't clear payroll checks on goodwill. This is Amazon though, so who knows how they think about it.

Our school district had a transportation manager who refused to buy electric school buses because the bottom line was they were more expensive to own and operate. He considered initial purchase costs, ongoing maintenance, the need to install and maintain a charging infrastructure for a fleet of buses, and the depreciation on the battery and replacement costs for same. He was replaced shortly thereafter by someone who promoted the party line. A referendum for a property tax increase was on the ballot within a couple of years. It passed of course, because "schools"

"Amazon bought the naming rights to the arena in June 2020, dedicating the arena name to bringing attention to climate change." - https://en.wikipedia.org/wiki/Climate_Pledge_Arena

From the EV manufacturer’s perspective, I imagine it’s likely easier to convince 1 “transportation manager” in your example to buy 50 school busses (single purpose vehicles) compared to convincing 50 consumers to buy 50 EVs each user with different driving patterns/needs. Likely also easier and cheaper to do maintenance on a 50 school bus fleet compared to maintenance for 50 individual consumer cars.

The business case can be well defined as expressed in your example, which makes solving for that problem easier. Selling en mass to the general public is a harder feat for an EV startup IMO

Tesla took the route of selling low volume luxury and then went down market. Seems like Rivian is taking a strategy of special purpose fleet vehicles (and I imagine eventually broadening their focus to consumers) - both are pretty smart business approaches to growth IMO

At the same time, anyone buying a fleet is expecting a discount for bulk. So the margin is thinner.

It is, for example, how airplanes have been so competitive despite there effectively being a duopoly. Pretty much no one is buying an airliner at list price.

School buses are a niche where everyone is buying about the amount of the same product from the same manufactures. You don't really get bulk discounts other than the same discount everyone else gets.

I don't think anyone buys them, period. They lease them. When the lease is up, they are re-leased or maybe sold to second-hand leasing companies or second-tier airlines, and ultimately they end up as freighters or scrapped.

The last of the DC-10s/MD-11s only recently ended service with FedEx. They may actually still be flying some MD-11s.

Buying is fairly common; Indigo just placed the largest ever single purchase of airplanes (500) https://www.reuters.com/business/aerospace-defense/airbus-wi...

If you’re leasing it’s not for free, the lessor is charging you a premium to make a profit over what they’re spending, so it still makes sense to own, particularly for large, heavily utilized fleets that don’t scale down much.

Yep, we used to deliver milk door to door daily in the UK with a fully electric fleet 50 years ago.

Could you say more about this?!

Wikipedia covers it reasonably well - https://en.m.wikipedia.org/wiki/Milk_float

There is a small "Powered By Rivian" badge on each side by the driver and "passenger" doors, if I remember right. The video walkthrough mentioned in the Ars comments calls it out specifically, and I was able to see it when stopped next to one at a stoplight.

I watched a video [0] (that I think I found here, maybe Reddit) months ago about these trucks (a driver did a full “review” of it), it was pretty freaking cool. I started watching it thinking “I’ll only watch a minute or two” and got so engrossed I watched the whole thing.

[0] https://youtu.be/3TFz1xqp3us

i saw the same video! the thing i was thinking about is, event though Amazon is a huge investor, the specs are so well designed for this kind of work that they could start selling to UPS, USPS, FedEx, etc.

Hey friend! Thanks for the share, super cool watch.

Hope you're doing well :)

Doug DeMuro did a rather comprehensive walkthrough as well: https://www.youtube.com/watch?v=3CWCqJl0BEs

Annoyingly, they didn't fix my main UX gripe with delivery vehicles, which is that they use the same lights for the hazards and turn signals. It's quite common for the right half of a stopped van to be obscured, which makes it impossible to tell the difference between "hazards on, delivering a package" and "left blinker on, about to cut in front of you".

> my main UX gripe with delivery vehicles, which is that they use the same lights for the hazards and turn signals

I don't know this for a fact, but it could very well be mandated by vehicle safety standards regulations. Of course, it wouldn't be very difficult to add extra lights like those you would see on the back of a tow truck or construction vehicle, but again that may be dictated by law.

I believe that extra turn indicators are allowed. Ambulances have a similar issue where they have dozens of lights but clear signaling of turns is important. A common solution is to add yellow arrows [1]. If I were designing a delivery vehicle I would include turn arrows that are not activated by the hazards.

https://www.fedsig.com/sites/default/files/products/Ambu.jpg

This gets into the absurdity of North American turn signal rules/regulations that allow brake lights as turn signals.

A relevant video by Technology Connections talking about said absurdity: https://www.youtube.com/watch?v=O1lZ9n2bxWA

This is the EXACT same absurd UX issue for public transit buses making stops. And even more absurd? It's illegal (in Ontario at least) to not yield to the turn signal. The issue is that with an obscured curb-side light, you can't tell when a the road-side light goes from 'stopped' to 're-entering traffic lane'. I have no idea why we can't add additional informational signalling to these special-case vehicles.

I thought it was illegal to drive past stopped busses in most of North America, no matter how carefully you drive or how many lanes are separating you? If it is there and has any kinds of light on, everyone else on the road must simply halt?

That rule is specific to school busses with the red STOP lights flashing, and you’re allowed to pass the other way if there’s a physical divider like a median in the road.

Commuter busses do not have this rule

> "left blinker on, about to cut in front of you".

You can't fix stupid. If the delivery guy is cutting in front of you without looking in their mirror, it's on them. Introducing a new set of lights will further confuse people who aren't delivery drivers and have no idea what those lights mean.

I think a flashing left arrow, à la ambulances, would get the message across.

https://www.fedsig.com/sites/default/files/products/Ambu.jpg

I see lots of these Rivian vans when I drive to work through a light industrial section near major Boeing facilities. This kind of local, stop and go delivery service is a perfect use case for EVs. Very similar to postal delivery which also cries out for EVs.

Wait... I think I might know the exact place you are talking about haha. Did Amazon move in in the past couple years?

This is in the SoDo area south of Seattle. I'm guessing that Amazon has a fulfillment center somewhere around there due to the density of Amazon vans I see on that road.

Oh nevermind then... they have another fulfillment center neighboring a Boeing facility outside of Washington, coincidentally.

> Very similar to postal delivery which also cries out for EVs.

USPS is very much on the same page: https://about.usps.com/newsroom/national-releases/2022/1220-...

I particularly like the job they did with forward visibility on these. Much like USPS trucks; minimizing blind spots was an explicit goal. I wish all trucks and SUVs and such in my city were like that, with either low front ends or cab-over-engine designs with good visibility for pedestrians and cyclists, especially little ones.

Are you talking about the same step van clone that requires a touchscreen to operate the cargo door?

What does that have to do with visibility?

There's probably only one Amazon Rivian van model, so if that's a Rivian Amazon van, then yes.

It's designed to be modular (different lengths especially). There's a smaller variant for European roads that was just in the news: https://stories.rivian.com/amazon-announces-rivian-edv-europ...

I would hope that drivers aren’t messing around in the cargo hold while driving - that sounds far more distracting than a touch screen control.

The touchscreen is in the flow as it is a digitally controlled door that ties position to gear selection for safety. There is method to the madness. I also wonder if they have plans for versions that are very different and thus putting it in the screen helps with reducing variance in the cab (no need for a second button, etc).

> The touchscreen is in the flow as it is a digitally controlled door that ties position to gear selection for safety.

All street legal vehicles have had simple, reliable interlocks with the shifter knob and hand brake for over 40 years with a single 12 volt lead and now you're talking software-based "flow" for safety critical functions as if this is some sort of novel improvement that should impress everyone. I'm left speechless. This is the kind of absolute madness only SV could appreciate.

UPS package cars and USPS mail trucks are caveman-simple for a reason. In fact UPS is just now getting air con for the first time in 2023.

> UPS package cars and USPS mail trucks are caveman-simple for a reason. In fact UPS is just now getting air con for the first time in 2023.

Is it because UPS is cheap and has an exploitable workforce?

UPS drivers are extremely unionized. They are hardly exploitable. Many are making six figures (with overtime).

And it looks like they're about to go on strike.

Yes, and the same van that has automated cameras watching the driver at all times, making sure they are paying attention while driving.

My milkman (London) has been driving an electic float https://en.wikipedia.org/wiki/Milk_float since forever.

This is not reinventing the wheel, more like rediscovering the wheel we have actually had for 100 years.

Moving what's basically an ultralight carriage at 20mph with some milk in it is a very different problem than moving a delivery van full of packages that goes highway speeds.

The milk float replaced one horse. The internal combustion engine on a delivery vehicle represents 160 horses.

You didnt need to have a range of 200km, and cruise at 70mph to get between number 14 and number 27 smith street and 2,000kg of payload. As for the payload - 4,000 pints of milk seems a suitable load for a milkman to deliver before morning tea time.

Interesting. I had no idea these were EV or Rivian, I've been seeing them in my area for months now. Seems to be most Amazon deliveries in my area (North NJ) are using them now, probably 80% these, 10% private vehicles and 10% of the larger UPS/FedEx-sized trucks.

Love the design (rear lighting, doors, dashboard [1], all the very functional handles and space design for easy movement). It's always funny how it takes electric vehicles for car companies to take risks or do something different.

It's much larger [2] than it seems from their pictures

Duug Demuro gave a full tour of the vehicle https://youtu.be/3CWCqJl0BEs

[1] https://topelectricsuv.com/wp-content/uploads/2022/05/Rivian...

[2] https://assets.aboutamazon.com/dims4/default/64815c9/2147483...

The 'something different' is a large battery to take power from that has no risk of outing the vehicle if you use it too much.

I hate statistics like these. Are these the number of packages delivered to a few neighborhoods in a few months? Or the entirety of Amazon's load worldwide for a year? How significant is 150M packages exactly?

Instead they could have shared – % of active delivery vans that are Rivian, % of the total package load handled by them, % of all miles driven, emissions reduced...anything more meaningful than just one large number without surrounding context.

They are called vanity measures for a reason. In this case I think it's appropriate since the source is Amazon PR looking for favourable coverage.

Amazon shipped 4.75 billion packages in 2021 in the US alone [0]. So it's obvious why they didn't talk about percentages.

[0] https://www.statista.com/statistics/1178979/amazon-logistics...

Yeah, and the next question is, is the numerator growing faster than the denominator?

Not exactly related, but I preordered a Rivian SUV and waited in line for about 2 years. By the time my turn finally came up, lots of other manufacturers had announced plans to release off-road or towing-capable EVs, many at lower price points, so I ended up cancelling. I really hope Rivian succeeds and I'm excited to learn more about their second vehicle line, but anyone purchasing right now needs to have a true early adopter mindset, because they're paying a huge premium to get a Rivian today instead of a similar car for half the price in a year or two.

Also, I have to admit, as much as I love what this company is doing, both the R1 line and the Amazon van look totally cartoonish to me.

the Amazon van look totally cartoonish to me

So would the ubiquitous Grumman LLV, if you hadn't grown up with it.

Early adopter, yeah, but hardly the same kind of “1.0” an early Tesla was. I’ve had my R1T for over a year and it is without a shadow of doubt my favorite car I’ve ever owned, and that’s coming from a die hard V8-loving car dork. It’s just SO good at everything. They really hit it out of the park.

Considering that they do more than 5 billion deliveries a year... it's awesome to see Amazon moving towards greener technologies as these are the kinds of moves that can make a difference (large scale corporate), but there is still a ways to go!

I had no idea these were produced by Rivian. I really enjoy the design of these trucks and I always smile when I see one on the street.

different title now ("Amazon has 5,000+ Rivian EV delivery vans on the road"), but these numbers seem... infinitesimally small.

maybe it's a bigger deal because the first Rivian trucks just hit the road a couple of days ago when i thought they first hit the road a couple of years ago?

So irrelevant metric to judge the impact of vans by number of packages delivered.

It's relevant if X rivian vans deliver more than x sprinter vans on average, but without the comparison, it is irrelevant.

these are cool

Reply on news.ycombinator.com