Comment by barbegal | Hacker News Reader

Comment by barbegal | Hacker News Reader

Parent Full thread

It seems they really got out of their depth with this project. There are a lot of photos and videos of aesthetic things that (arguably) Dyson is good at such as the steering wheel user interface, the shaping of the wheel rims and the colour palettes for the paint job. But there is very little evidence of hard engineering, only one grainy video of the prototype car actually running (the other videos appearing to be the car running are definitely good quality renders).

As Elon Musk put it "For cars it's maybe 100 times harder to design the manufacturing system than the car itself." and I think as soon as it dawned on them the scale of the manufacturing process to build this car and all the precisely engineered systems they got cold feet. Yes conceptually a car is similar to a vacuum cleaner, there's a motor, a battery, a control system, a chassis and bodywork but they are in a completely different league in terms of the scale and engineering. A car needs to be able to run safely without faults for 8 hours a day 365 days a year whereas Dyson vacuum cleaners tend to have a lot of reliability problems even when using for an hour each week.

They also struggle with the culture of software.

From what I've heard from employees, software is looked down on by management and is not really something they 'get'. Things like app connectivity are after thoughts and not core to their products.

A good chunk of what makes a Tesla unique is the radically different software and digital user experience. Dyson (the company) are not ready to embrace the fast moving culture that enables that.

> From what I've heard from employees, software is looked down on by management and is not really something they 'get'. Things like app connectivity are after thoughts and not core to their products.

That is a surprise - of all the IoT devices I've owned Dyson's products have been the most rock solid and polished. Their mobile apps (on iOS at least) have been absolutely flawless, never requiring me to sign in again and again or having connectivity issues with WiFi. Their appliances also don't seem to do any funny stuff over DNS unlike some of the flakier stuff I've owned (one of which was making 20k DNS requests a day to a non-existent diagnostic server address on the local subnet).

> Things like app connectivity are after thoughts and not core to their products.

Good!

God help us the day something like a fan or a vacuum cleaner needs "app connectivity".

A good example of intelligent appliances would be like my Samsung washing machine... It had an error, so I called their support line. They asked me to hold down two buttons simultaneously, then put my phone near the door. It let out a 10s or so digital song, the kind that modems did in the 90s - that gave them an entire diagnostic, age, model etc, and they sent someone around as they knew I couldn't fix it. Pretty amazing stuff.

A counter example of an unintelligent washing machine:

My Whirlpool emitted a diagnostic puddle of water underneath it. I Googled the seal, ordered it from an independent supplier, and installed it myself.

It is now on its 25th year of use with us, and cost less than $400 when new.

While I, too, am using an ancient washing machine in my home, it's not exactly an apples to apples comparison. New washing machines use far less water and electricity than their older counterparts. They also remove more water during their spin cycle, reducing the amount of drying necessary. And if your dryer uses natural gas (as mine does), a newer washing machine will reduce your use of natural gas as well.

What kills me is that every feature you mention can be done without needing an app, an lcd, or the cloud, and the same thing applies to electric cars.

How much cheaper would an electric car like a tesla be if it didn't have autopilot, electronic keys, and LCDs everywhere? How much more secure would it be without an e-key to unlock it? I hope when I'm ready to buy an electric car I can find something like that.

I believe the Tesla Models 3 and Y's LCD screens were a cost cutting measure -- one LCD replaces dozens of smaller screens, switches, buttons, and indicators, as well as the PCBs and metal chassis to support them and the molded plastic to hide the PCBs and chassis.

It's probably also less engineering effort to put a set of controls on the screen than to design the corresponding physical controls.

I’m willing to bet the LCD screens and fancy hardware are a small part of the vehicle cost, esp. when compared to the batteries.

All software is expensive up front but near zero cost to deploy to each additional car.

At some number of cars sold the software costs becomes a rounding error.

Except when that software needs more than a microcontroller to run. The computers Tesla throws into the cars are not super cheap.

Car manufacturer tries to reduce penny costs for each parts literally. Good SoC/LCD are $$$.

Meanwhile, software is costs at once so traditional car manufactures uses cheaper chip and force programmers to write efficient code.

The physical hardware might be a drop in the ocean on a $100,000 car, but the software engineering for all the features they enable cannot be cheap.

I'll be interesting if the gadgetness continues once the cost to build an EV drops below a basic gasoline powered car.

My commute is 8 miles of which 6 is on a freeway. Doesn't feel like a burden. Unlike say a tech bro doing his 50 mile commute down 101. So I'm dubious an autopilot would be a big win from me.

I'm imagining your Whirlpool singing a happy diagnostic song while it wees on your floor

> Pretty amazing stuff.

Pretty terrible stuff. If it uses proprietary tones as a diagnostic code you can't diagnose and fix it out yourself.

A little bit of me thinks that smells like a smoke and mirrors show as an aid to selling. Last time we got a washing machine repaired was over decade ago and it didn't last long. Pretty quickly realized that replacing worked out cheaper especially if you don't throw money away on stuff like calgon every wash. Happy to buy cheaper dumb washing machines.

That's the opposite of a smart appliance to be honest.

You can use minimodem[1] to scream a file across the room from one computers speakers to another computers modem.

It's a cute trick but requires very little in the way of computation, and is a far cry from the bluetooth infested dumpster fire that is found in the modern app-controlled smart appliance offering.

It's also not significantly an improvement from giving your products clearly labeled short model names and having a two digit code on a 10c LED screen, preferably with a corrosponding entry in the manual. I'd happily give it up in exchange for a washing machine with minimal electronics, so that it will last 20 years and can be fixed with duct tape and epoxy.

[1]http://www.whence.com/minimodem/

I can see that being useful if the app said, "Replace O-ring AE-35. Press ORDER to order this part, INSTRUCTIONS for service information and diagrams, or SERVICE to schedule a service appointment."

But my guess is that pretty much every possible outcome of the diagnostic process is "Make sure an adult is home between noon and 6 PM next Tuesday to let the repair technician in." It's safe to say that empowering the consumer is pretty far down the list at Samsung's R&D department.

A web of sensors enabling diagnostics (few accelerometers, temp sensors, even more water-presence ones, maye even strain gauge, cabling, microcontroller with enough analog IO, intelligent motor controller, both BOM and labor) would offset any imaginable savings. At the end of the day the outcome is always "we will send someone". Sounds like a dog and pony show, expensive scam.

I agree it probably wouldn't result in a lower repair bill, but that's only one cost of appliance repair. The other major one, which isn't a big issue at the moment, but usually is... you have to be home when they come to service it.

Not everyone can work from home, so the shorter the visit the better. If the diagnostics are done remotely, they can arrive with the right parts and do the job more quickly since they already have a headstart from the diagnostics. Additionally, they should be able to complete the job in only one visit since they should already have the parts they need and not have to order them.

LG has a similar system too, and it makes sense for everyone. They save money on warranty repairs since their labor should be lower. Consumers save time on repairs for the same reason. They also get to drive repair work to the service company of their choice, which has benefits for them and has some pros and cons for the consumer.

The downside to "intelligent" appliances is that the added complexity of that intelligence is not free, it adds a lot of new potential failure modes, particularly when they are connected to a network. The benefits to adding it have to be significant to be worth it, at least for me to consider purchasing a smart appliance over a dumb one.

A "connected" fan definitely has its uses, especially if it can sense if I'm in the room or not.

"Connected" vacuums are already here, but they're called Roomba, not Dyson.

Dyson Eye 360: https://www.dyson.co.uk/robot-vacuums/dyson-360-heurist-over...

Although the connectivity with their app is a bit rubbish half the time!

True. Although, I have a secondhand fan / air purifier of theirs, and in theory the smart features of that are ideal - setting timers, voice control, etc.

But it's crippled by poor UI, only supporting Alexa, and an arbitrary limit of how many schedules you can set, etc.

We already have app connectivity for the robot vacuum cleaners.

sounds like you're experiencing technophobia

>A car needs to be able to run safely without faults for 8 hours a day 365 days a year whereas Dyson vacuum cleaners tend to have a lot of reliability problems even when using for an hour each week.

To be fair, we're talking about a British car here. The bar is lower.

Dyson should have gone with another manufacturer's EV platform for their first few cars, like Volkswagen's MEB platform:

https://jalopnik.com/volkswagen-wants-other-automakers-to-us...

https://jalopnik.com/the-fascinating-engineering-behind-vws-...

It wouldn't have been a Dyson from the ground up but it would have gotten their EV effort started at the lowest cost and they could have worked on their own in-house designs in the meantime.

If they had gone with MEB, what would they be bringing to the table? Dyson's bread & butter is electric motor design and you're suggesting they outsource that and instead build a shell around someone else's platform? Arguably that's their weakness.

Tesla started out with retrofitting their electric drive train into a Lotus Elise, that approach would have allowed Dyson to do what it does best.

> what would they be bringing to the table?

Brand and design.

So what would be the point?

The point would be to get their EV business started at the lowest cost.

MEB is a platform, not a car. There's plenty to do beyond the platform. Ford will be making an MEB car. Here are some of Volkswagen's planned MEB cars so far:

- VW ID.3: https://www.youtube.com/watch?v=IPjvgXWA78E

- VW ID.4: https://www.youtube.com/watch?v=9dh00utug10

- VW Space Vizzion: https://jalopnik.com/the-volkswagen-space-vizzion-brings-the...

- VW Buzz: https://www.youtube.com/watch?v=nfKYr9sgxnw

- SEAT el-Born: https://www.youtube.com/watch?v=Evr55WN5XWo

- Skoda Enyaq: https://www.youtube.com/watch?v=hLTBHGAv5aw

- Audi Q4 e-tron: https://www.youtube.com/watch?v=DiwevzHsCbU

MEB is flexible enough to work with many makes and models of car. MEB is exactly the sort of platform a company new to the car business would want to start with.

Or perhaps the opposite. They could have designed a motor and battery package which other manufacturers could build into their own designs.

Thus they would get a start into electric vehicles without needing to invest in the entire production process. From there move to a Dyson branded car manufactured by one of their partners (designed by Dyson in collaboration with their partner) and then, if worthwhile, step up to manufacturing the entire car themselves.

I'm not sure electric cars are as deep as motor driven cars and obviously Dyson had the manufacturing of electric motors down to a fine art. I am pretty gutted they couldn't get an economy of scale to make it work, even if they had sold 100 at 250k a pop I think it might have grown and they might have followed the tesla model from luxury down to affordability.

Electric cars are as deep as you make them. I've watched several youtube videos showing disassembling some Zoe components, it's really impressive. Some models use parts of the motor as an inverter during charging.

Dude! This has got to be the same with software -- with software it's 100x harder to design the process than the software or at least an order of magnitude wouldn't you say?

My first thought was cooking. It's far easier to make a great meal once than repeatably in a restaurant on an assembly line.

Reply on news.ycombinator.com