A Fuel-Efficient Big Rig From Walmart | Hacker News Reader

A Fuel-Efficient Big Rig From Walmart

wired.com

Here's another truck that is more conventional in design, yet doubles fuel-efficiency.

http://www.airflowtruck.com

It has full skirts, aero nose, and a trailer tail. The only place left to optimize is the tractor to trailer gap, responsible for about 25% of the aero loss in a regular truck.

If a company is going to get these vehicles to be market viable, it is Walmart. They are ruthless in terms of cost cutting and they would be the first to cut out "tech for tech sake" if it doesn't give them a ROI.

I've learned from programming that cost cutting is always the n-th step in creating a new product. What we are talking about here is pioneering into a new territory. Cost cutting isn't even a question right now.

WalMart has the Glass center who's whole purpose is technology for cutting cost. Its a thing they do.

Worked there for 9 years. Its funny. Outside of this area, it doesn't occur to most people that Walmart would have IT, but they've got more people in that building than some IT companies.

What kind of stuff do they do in there?

Ever read the comic Dilbert?

Okay, kidding aside, they have teams for different aspects of infrastructure, like building servers, network admin, database admin, security, etc.

Then they have areas dedicated to systems for item tracking, training, accounts receivable, accounts payable, benefits enrollment, etc. Within these areas, they have teams of project managers, programmers, QA, and 2nd level support.

Another large chunk of the building is dedicated to 1st level support and middle management.

There's R&D guys and there's technical experts who look at how the different systems fit together and help with strategy for the technical aspects of the big picture.

Then there's the walmart.com guys who are actually in California. And there's teams dedicated to the mobile sites for walmart.com and samsclub.com.

And there are meetings, lots of meetings. And process. Lots of process. To install any small change will probably take you 3 days if its an emergency. Otherwise, expect a week minimum. Projects in more recent years take upwards of 1 to 2 years to complete, using the waterfall methodology.

Is Walmart Labs / Kosmix part of that R&D group or is that something different?

Kosmix was acquired by Walmart in 2011, but I hadn't heard of them before your question. WalmartLabs is part of the R&D group that I mentioned.

Sam Walton talks a fair bit about tech in his autobiography. He says he had to be talked into all of it, but he thought it paid off.

One thing that impressed me all the time was that any small improvement, when done at Walmart's scale, meant big numbers. I don't remember the specifics, but when one guy figured out how to make the printed receipts use less paper, that was millions of dollars saved.

If Walmart can save a little bit per truck using these trucks, it'll be big savings.

This seems very much to be aimed at cost cutting. What new territory is being pioneered here?

I've always wondered about fuel efficient trucks. A number of 'design the future' efforts include them but they rarely make it to the road. The industry doesn't seem to care enough to make the investment.

I really think that self driving fuel efficient trucks on their own "truckways" would be a pretty good way to boost transport infrastructure quickly.

And self driving busses could run there too. Maybe rebuild the old fashioned legacy railways as modern truck-bus-ways.

Commercial vehicles are utilitarian above all. Efficient semi truck designs pop up all the time but they always fail because they sacrifice ease of service to achieve their gains in efficiency. If a Volvo semi breaks down while hauling one of Wal-Mart's JIT deliveries it can be repaired by any shop in hours. If a highly proprietary semi breaks down then it's hosed. Either Wal-Mart rush processes another truck to that store or they eat a lost delivery and make up the missing inventory over the next few trucks.

Repairing all the advanced, low production volume technology in that truck is a lot more expensive. Add the significant amount of money lost when JIT deliveries fall days behind and the expenses far outweigh the fuel savings.

In some ways, you describe freight rail. Freight rail in the United States is great. We use it a lot more than Europe (probably about 4x more) and the fuel efficiency gains are huge for freight rail. It isn't self-driving, but it is on dedicated tracks with minimal people needed given the long length that trains can be.

I think this is the reason why the industry doesn't seem to care: they do care, but they've already invested in freight rail which is even more efficient. It's going to be hard to beat freight rail's efficiency and at what cost? Trucking does have greater flexibility and is needed to do shorter distances and last-mile stuff, but we already have fuel-efficient trucks on their own truckways: it's freight rail.

Yes, and the trucking industry stopped fighting freight rail years ago and now makes use of it.

Trucks are more flexible, sure, but to get that flexibility they trade off operating efficiency. This is why many long haul trucking companies such as J.B. Hunt [1] and even UPS [2] now make use of intermodal shipping. [3]

It's not commonly known that packages sent via UPS Ground frequently travel on trains. As a general rule, UPS Ground packages traveling long distances across the U.S. are shipped by train. [4]

[1] http://en.wikipedia.org/wiki/J._B._Hunt

[2] http://www.ups-scs.com/transportation/rail.html

[3] http://en.wikipedia.org/wiki/Intermodal_freight_transport

[4] http://compass.ups.com/processes-of-package-in-transit/

An enormous part of the industry consists of contracting owner/operators hauling client (third-party) trailers. That is true even of many of the "fleet-looking" tractors you see on the road; the "fleet" is more of a contracting agency for nominally independent operators than an actual fleet, and the livery just means that the trucker contracts with the agency (the truck itself is either driver-owned or is on a rent-to-own sort of arrangement that makes the "no payments until 2075!!!" furniture guys look good). Since most of the benefits (streamling, regenerative braking, etc.) can only occur in paired tractor/trailer rigs, it's only those outfits that own both ends that can implement much, and then only if they have no retirement sales in mind.

From the comments, a better YouTube video explaining more: https://www.youtube.com/watch?v=NER9X4_gtYk

OT: Native english speakers, got some questions. I'm constantly trying to improve my english online, so ..

3:42 "The emissions is extremely low" -> Is low? Are low? Why is?

3:46 "Diesel emissions fuel" -> Is that the canonical way of spelling 'Diesel'? I mean, that's more or less exactly the way I'd spell it, but .. I'm surprised. Usually english doesn't quite like 'i' as in 'ee'. I know that this is a German name after all, but .. I have a number of peers and friends that wouldn't be able to map 'Morpheus' (Matrix), 'Zeus', 'Euler' etc. to the German pronounciation, i.e. you'd get a 'Err.. WHAT?'. Diesel made it 1:1? Or is that debatable and there are people that pronounce it differently?

1) That should be 'are'. It's just a mistake.

2) Diesel is named after the inventor of that type of engine, Rudi Diesel. We still call Euler 'oy-ler' where the expected pronunciation may be 'you-ler'. I don't see any inconsistencies there :-) Morpheus and Zeus aren't German names, so you wouldn't expect English speakers to pronounce them in a German style, right?

re: Diesel emissions fuel. He may be referring to Diesel exhaust fluid, basically liquid urea that is injected after combustion to cute down on emissions.

http://en.wikipedia.org/wiki/Diesel_exhaust_fluid

At 200000 miles a year and 6mpg average fuel efficiency, a 1 mpg improvement is $25k in savings. I'm surprised this hasn't happened sooner.

Actually, that sort of logic is probably exactly why it hasn't happened. The marginal increase in costs for this is almost certainly much much higher than $25k.

Yeah, but the $25k is recurring so you would have to capitalize it to compare it to the one-time cost of the new truck.

Its not just capitalization of the up-front cost, its also the delta in maintenance cost.

A cab and trailer is probably already something like $200-300k already, so I could see this sort of truck easily being more than $100k more expensive just in the upfront cost.

But trucks tend to last 10+ years. Not sure what the lifetime of trailers are though.

Turbine powered battery electric drive train that can burn multiple fuels. It seems like that would be a nice combination in a car that you could drive long distances on cheap fuel without worrying about recharge stations and still get electric motor torque and regenerative breaking.

And compressed air regenerative breaking.

Since the siding is no longer metal, does that mean that it no longer acts a a Faraday cage, and will be susceptible to damage by lightning?

...so long as they remember to only turn right (UPS style) then this should be fuel efficient.

In reality though this truck is not currently or likely to be economically viable. In Europe Mercedes have had a problem selling their latest and greatest 'eco' truck as it costs more and the fuel savings just don't cover the added up-front expense.

If carbon emissions were priced, then maybe these kinds of technologies would be more viable.

In many European countries, they are, as part of a fuel tax: https://en.wikipedia.org/wiki/Carbon_tax#Europe

While fuel is more expensive in Europe, their trucks are actually less aerodynamic, with big flat fronts [1]. This is because truck lengths are strictly limited, so a shorter cab means you can run a longer trailer. Even with their high fuel costs the gains from the longer trailer outweigh the savings from better aerodynamics.

[1] http://www.emercedesbenz.com/Images/May08/15_Mercedes_Benz_C...

> If carbon emissions were priced

They are. Fuel isn't free.

Surcharging carbon dioxide would affect the choice of fuel, but would have no effect on aerodynamic efficiency since that applies to all fuels equally.

Big rigs are not concerned with turning right vs left for efficiency. They rarely need to turn at all, because they do long distance hauls (either between states or cities).

Except when they're navigating to/from the highway and their destination within the city, picking up and delivering loads, and stopping at a truck stop for fuel, food, or rest.

The above is just for OTR (over the road, i.e. long haul), and you've got regional drivers, many who spend all day making deliveries within a much, much smaller geographical area.

It is likely more efficient for trucks to only turn left. Turning right often takes more time than turning left, as a right turn can often involve waiting for traffic to move out of your way. Back when I drove trucks, any time I had to drive in SF, I purposely chose left turning routes for this reason.

The reason fuel efficient trucks are so much more expensive is because they are bought by companies with accountants and finance departments instead of guilt-ridden liberals. They have to actually demonstrate ROI in order to sell well. With the margin padding that manufacturers and dealers expect, that ROI is often washed away. With a company like WalMart doing the design, with contract manufacturing, they likely could make a decent return.

Thanks for your comment and perspective. My quip regarding UPS was a carry over from a thread that was on here yesterday with not a lot of thought beyond that! (Allegedly for UPS it is more efficient for them to do routes with no right turns.)

It figures that big trucks prefer right turns but until you pointed it out here I had not put two and two together, despite anecdotal evidence I see on a daily basis. This is actually useful knowledge as I share the road with the big trucks. I ride one of those vehicles where I am the crumple zone (a bicycle) in London where a lot of cyclists get killed down inside the left-hand side of left-turning trucks (we drive on the left and call 'trucks' 'lorries'). There is a lot of consternation about this with truck drivers tarred with the same brush quite unfairly (in my opinion they are professional drivers, cyclists, nah).

A little bit of understanding could help. In all the debate I have seen I have never seen it mentioned how left turns (in the UK) are particularly tricky for long vehicles as they need to use the on-coming traffic lane of the road they are turning into or the road they are on. This little detail needs to be widely understood by cyclists, it is not as if lorry drivers are being deliberately retarded and un-caring, a left turn is something very difficult for them and well beyond what a mere cyclist is capable of.

I saw this when it aired in the 80s: https://www.youtube.com/watch?v=9ndY8g0cuEI

Walmart could be saving lots of money now if they wanted to, and without any special technology.

That looks like a normal truck on the road today. Isn't that what they have now?

Walmart's newer fleet already tends to be trucks and trailers with side skirts and other aerodynamic features, plus tend to not be of the cabover design which pretty much mandates a brick wall of a front end that rams into the air around it. Most prototypes, like what this truck is, tend to be like high fashion designs. Yes, they look ridiculous and extreme and weird, but the aspects, the bits and pieces of the designs tend to show up in the finished product.

A truck like this lets them see with one rig how much is saved by the aspects they want to explore -- a new cab design, even more radical aerodynamics, a new powertrain, new materials for the trailer -- altogether, and do tests to see which areas are worth further exploration in a finished product. Plus, it's flashy and eye-catching, which is useful for the departments that work on these sorts of things. It lets them say, "hey, we make neat things, and get people talking about walmart, keep funding us."

I can't imagine how much those 53-foot carbon fiber panels cost.

...or how much the energy bill is for the Airbus sized autoclave needed to make them.

solar oven would be great to bake these

I am very surprised by the turbine engine. They are not know for being fuel efficient.

They are not efficient in a car which is constantly stopping and starting. If they can run at a constant speed (e.g. driving a generator, propelling an aircraft in cruise) they can be much more efficient.

"Arise, Rodimus Prime!"

It looks like a train and a sports car had a baby

I am a little skeptical that air resistance is a huge factor for an 80,000 pound fully loaded semi. Probably most friction losses are due to rolling resistance. As it’s a sunday and I’m daydreaming anyway, I might as well try some back of the napkin calculations:

http://www.engineeringtoolbox.com/drag-coefficient-d_627.htm... http://www.engineeringtoolbox.com/rolling-friction-resistanc... https://en.wikipedia.org/wiki/Semi-trailer_truck

Before I start, I don’t know if the Engineering Toolbox answer is wrong, because I get 217.5 N for (0.29 * 0.5 * 1.2 * ((90 * 1000)/3600)^2) * 2. At least it’s close to the 181 N they claim. So my results could be off but I’m hopeful they’re within the correct order of magnitude..

If we assume a drag coefficient of 1 for a flat plate representing the front of the truck, with most losses coming from form drag, A semi with a frontal area of 2.5x4 meters at 105 km/hr (65 mph or 29 m/s) would feel a force of:

Fd = 1.0 1/2 (1.2 kg/m3) ((105 km/h) (1000 m/km) / (3600 s/h))2 (10 m2) = 5104 N = 520 kg = 1147 lb

If we assume the high drag coefficient of 0.01 for truck tires, the rolling resistance of a 36287 kg (80000 lb) truck (idealized as independent of velocity) is:

Fr = 0.01 (36287 kg) (9.81 m/s2) = 3560 N = 363 kg = 800 lb

Power P = F * v so these are:

Air resistance power = 520 * 29 = 15080 W = 20.22 hp Rolling resistance power = 363 * 29 = 10527 W = 14.12 hp

So on flat ground, amazingly it only takes about 26.1 kW (35 hp) to keep a semi moving at 105 km/hr (65 mph)! This makes sense to me because a human can pull a semi at about 5 mph. Power goes up roughly by the square of velocity so a human with a power of 300 W would need 13 * 13 = 50.7 kW (68 hp) to go 65 mph. We’ve seen here that roughly half the power goes to rolling resistance which stays roughly constant with velocity, so 35 hp is conceivable.

As a sanity check, climbing a 5% grade at 105 km/h is a climb rate of 5.25 km/h or 1.46 m/s. The power required to lift the mass of the truck is:

Power P = F * v = m * g * v = 36287 * 9.8 * 1.46 = 519194 W = 696 hp

So it takes toughly 20 times more power to go up a 5% grade, and that’s why a typical semi truck engine with 600 hp can’t climb a 5% grade at 65 mph.

Semi trucks get 5.5 miles per gallon, so use about 12 gallons per hour. There are 38 kWh of energy in a gallon of diesel fuel, so that’s 456 kW burned in an hour, or a continuous power used of 456 kW (612 hp). If we compare that to the ideal of 35 hp needed, we can see that:

Engine to road efficiency of a semi truck = 26.1 kW/456 kw (or 35 hp/612 hp) = 5.7%

That’s really quite remarkably low and doesn’t surprise me, since semi truck technology has not changed much in 50 years. I read somewhere that cars are in the 8-15% efficiency range. That’s why it’s so trivial for Tesla to beat an internal combustion engine, because electric motors are 95% efficient vs. about 25% efficient for an internal combustion engine.

So the real increase in efficiency of a hybrid gas turbine/electric semi comes from the turbine, which has an efficiency of 40% vs about 25 or 30% for a reciprocating diesel engine. Improved aerodynamics and decreased rolling resistance of a carbon fiber trailer are mostly for show. Hybrid turbine/electric propulsion has been used on trains for decades and I’m kind of flabbergasted why it never took off in semis, or cars for that matter.

Haven't you noticed over the last 15 years or so that big trucks nowadays basically all have those cowls on top and they're a lot more rounded on the front.

Turbine systems scale up in efficiency as they get larger. Small turbines are not necessarily any more efficient than diesel engines.

> Hybrid turbine/electric propulsion has been used on trains for decades…

I'd love to see a citation for this statement—it's far more accurate to say that turbine/electric propulsion has been tried for decades. Since 1969 there have only been a few test units built. [1]

The largest fleet of gas turbine-electric locomotives was operated by the Union Pacific until 1969. "Fuel economy was poor" and the units were not considered a success. [2] These units are quite well-known in railroad circles.

If gas turbines were as good as you say they'd be in heavy use on railroads today. Instead the dominant motive power is electric for lines already under catenary and diesel/electric everywhere else.

[1] http://en.wikipedia.org/wiki/Gas_turbine-electric_locomotive

[2] http://en.wikipedia.org/wiki/Union_Pacific_GTELs

Ya true, I guess you are right. I thought trains used turbines because it would just make sense (there's no transmission, so the engines can run at one optimal speed). I'm thinking that maintenance costs/reliability issues make turbines impractical in most cases except aviation and power plants.

On that note, I've never been much of a fan of turbines anyway, because they are outrageously complex/expensive. I think a Tesla turbine would be fantastic for applications like this though. For one thing, the disks could be cast out of ceramic. There must be some reason why they aren't being used (especially in hybrid electric systems like this) but I can't think of one.

Reply on news.ycombinator.com