[TorqueyTechie]

Hi all,

I was talking to a colleague of mine regarding cars and an interesting point came up regarding the state of the ICE display unit. He mentioned that the UI display in his 2013 honda is not good. This started me to look up more on that and I found an interesting thread which mentioned about the same in premium car segment. Below is the URL :

http://www.teehanlax.com/blog/the-state-of-in-car-ux/

Looking at the display of most of the premium cars(ferraris, Porsche's) I got reminded of the old touchscreen display phones that we used to get.

Is it so costly to develop an UI which matches to the price and segment where it is being targeted?Apple and Google have forayed into this recently with their products but still I believe this is one area where development can be still done.

Any thoughts from fellow bhpians?

[Jeroen]

Well, what is a good UI is probably very much in the eye on the beholder.

I recently got my wife an iPhone and she doesn't like it. Claims it is completely counter intuitive. I pointed out that in the Netherlands 50% of the people over 50 years of age all have iPhone. That makes the Dutch very smart or the Apple UI very easy to understand. She doesnt agree.

Anyway, for me in order to like a car, the dashboard is a very important part of the total package. if I dont like the dashboard, the look and feel and how it works, I'll never buy the car.

Years ago I got close to getting a Peugeot 406 coupe. Beautifull, gorgeous car, designed by Pininfarini, but it had the most drab dash you could ever imagine. So I did not go for it.

I like complex/busy dashboards. The more dials, knobs and lever the happier I am. I love the old British style dashboards with switches and stuff thrown randomly about.

Having said that, I do appreciate the modern dashboard, such as you will find in many high end cars today as well. Very simple, sleek, and usually some sort of menu driven list of endless options.

When I get something new I will actually read the manual back to front before I start using it. That would go for my electrical tooth brush, smart phone and car or anything else. Got a pair of headsets the other day, read the manual first.

To me, it's just part of the enjoyment I get out of a new car. Sit in the car, go page by page through the manual and check out everything in the manual how it works in real life.

I remember getting into the new Jaguar XJ a few years back in the UK and the sales guy took more then two hours to explain the menu and the various settings. Great stuff, my wife absolutely hated it.

My neigbour back in the Netherlands used to be in charge of the Mercedes workshop near where we lived in The Hague. He got some health problems and he wasnt allowed to work full time and had to take it easy. He was very good with customers and they put him in charge of the customer familiarisation with their new Mercedes. So he would sit next to the customer in their brand new car and explain how it all worked. He told me he could not do it under two hours and that was on the base models. Anything more fancy he would need at least half a day.

He was always very upset with customers that just pressed start, slammed the gear into D and drove off without bothering with as much as adjusting the mirrors.

Let me share another experience. I'm no expert, but I have been driving and working and reading about cars just about all me life. I must have driven hundreds and hundreds of different cars with no problems. During our stay in the USA I travelled a lot and everywhere I rented cars. Never a problem. That was until Hertz gave me a Prius. At first I had no idea, just got in and got going. Then, I had to pull over somewhere to make a call or look on the map whatever. And I could not get this stupid car to move again!

I tried everything, there was no manual, there never is on a rental. So eventually I just called Hertz. They said, Oh you're in a Prius, happens all the time. Not sure what I had to do or what I did wrong, but it was something along the lines of put your foot on the brake, move this switch, put into gear etc. Anyway it went again. But as a UI it was utter crap. So that is my norm. If I cant figure out without a manual, and that would be rare, it is a particular bad design.

Not sure what makes a generic good UI. Do 50% of the people need to like it. 95%? i just dont know. Obviously, getting my wife to like any UI is a challenge to wich not a single design team in any industry has rissen. True Mission Impossible.

Takes all sorts.

Jeroen

[GTO]

Totally agreed. Worse still, in some OEM UIs, you can actually see the pixels!!!

I've always maintained that automotive ICE & entertainment is a huge growth opportunity for iOS & Android. Everyone has a car! Not too far away, I guess : Related Link

From our Mercedes E-Class Review

Center screen fonts are very 90’s. Easy to read albeit lacking in finesse. Graphical interface can't even compare with windows 3.11!!:

[ampere]

Originally Posted by GTO

Totally agreed. Worse still, in some OEM UIs, you can actually see the pixels!!!

I think the problem with outdated infotainment is that you need semiconductors which qualify for Automotive grade conditions; which are far more stringent than mobile and even industrial grade. The same processor which can power a cellphone cannot just be added to the car (even if its for infotainment and not critical system control). It needs to be designed with a much higher reliability requirement and very low FIT Rate (Failures in Time). Even if such processors end up in the market, thanks to high margins, they might only make it to high end cars.

However things ARE looking up. With the current explosion of the Automotive sector in third world countries, these requirements for semiconductor designs are also becoming mainstream.

[TorqueyTechie]

Jeroen, I do agree with you that looks are subjective. but my point was that for some of the cars which costs upwards of 3 million, atleast a better UI would go in a great way to enhance the interior look of the car.

OT : I do agree with your wife on iPhone. I dont like them either but I may be biased since I am an android user.

GTO, exactly my thoughts. My friend has a 2012 Mercedes C class and the UI looks awful. The one that I have liked is from 2013 honda accord which was slightly better with the look and feel.

ampere, I partially agree with you on that. The premium cars are all built after a proper research and analysis. If little amount of the money is used up for enhancing this aspect then it would help build up the customer satisfaction of the brand IMO.

[Jeroen]

Originally Posted by ampere

I think the problem with outdated infotainment is that you need to semiconductors which qualify Automotive grade conditions; which are far more stringent than mobile and even industrial grade. The same processor which can power a cell phone cannot just be added to the car (even if its for infotainment and not critical system control). It needs to be designed a with a much higher reliability requirement and very low FIT Rate (Failures in Time). So in the event even if, such processors end up in the market, thanks to high margins they only make it to high end cars.

However things ARE looking up. With the current explosion of Automotive sector in third world countries, these requirements for the semiconductor designs are also becoming mainstream.

Are you sure? I don't think its the processor that is the problem. Plenty of powerful processors around that would do fine in cars. Even on mobile phones to date the bigger problem tends to be the screen. Getting very tough, resilient high resolution screens is a problem, although much progress has been made. But even to date, you drop any smart phone, you are likely to damage the screen, not the processors. There are hundreds of thousand of planes out there with very advance glass cockpits with truly amazing screen technology and UI. So, i don't think process power or the electronics can be a big thing. Price of electronics is just a volume thing.

Here s another thought, even today smart phone designer are struggling on designing UI, although some have come a long way, e.g. Apple, Samsung to name a few. But you would not believe the amount of research went into that over many yearsNot just technical research, but also human behavior research. I just wonder if car manufacturer have the capabilities to think this sort of stuff through from the drivers perceptive. Its part science, part art, part magic and gut feeling.

If anything this must be a relative new area of expertise and competence for many car manufacturers I would think. But then again, I don't really know. I know several individuals who are car designers, figuring out shape, form of the car, engine, but I have yet to meet the first car manufacturer R&D guy specializing in UI, from a non technical point of view. Some members seem to have quite good insights into certain manufacturers. Would be interested to hear their thoughts?

Jeroen

[dj78]

The situation with the head units is going to change sooner than later. Almost all OEMs (lead by the luxury brands) are working on developing ecosystems that deliver apps via these head units à la apple and google. I think Volvo has the most mature offering at this time but GM, MB are also pursuing tis aggressively – This is going to be the next big channel for the OEMs to remain in touch with the customers after they have driven off and they don’t have to depend only on the dealerships. There are many other benefits to the OEMs – e.g. with appropriate connectivity options over the air software update for ECU preventing costly recalls..The problem is each one is developing their own closed world and hence the concept hasn't seen much adoption.

Google has seized the initiative and has recently launched the Open Automotive Alliance (http://www.openautoalliance.net/#about) - It intends to develop a version of the Android platform for automotive. So far Audi is the only luxury brand onboard but who knows it might take off just like the Open Mobile Alliance did and the rest is history.

Apple is also playing catch up - CarPlay is its gig.

Unfortunately the market is very fragmented right now with several OEM, supplier, Telecom Network Operator.. led ecosystems which is further complicated by the different regulatory requirements (related to driver distraction et. al) for different markets. Some consolidation will happen eventually and then the adoption rates will go up.

[ampere]

Originally Posted by Jeroen

Are you sure? I don't think its the processor that is the problem. Plenty of powerful processors around that would do fine in cars.

As I said in my earlier post, processors which power the cell phone can potentially find themselves on most of the low to mid end cars. But they have not been able to. Its not just the display alone.

The other reason is that the product life cycle of a phone/tablet is much smaller in time, as compared to a car. And once a processor is qualified for a car platform; that platform stays for a long time. Hence the revenue cycle time is much longer as compared to a mobile product.

Even on mobile phones to date the bigger problem tends to be the screen. Getting very tough, resilient high resolution screens is a problem, although much progress has been made. But even to date, you drop any smart phone, you are likely to damage the screen, not the processors. There are hundreds of thousand of planes out there with very advance glass cockpits with truly amazing screen technology and UI. So, i don't think process power or the electronics can be a big thing. Price of electronics is just a volume thing.

Screen is also an added problem. Thats why if you see industrial displays are not very high resolution but much more error resilient. (Your own words.)

As regards to processor, it was never the performance but always the reliability. Still there is a basic performance requirement to power a decent display, run videos etc.

Here s another thought, even today smart phone designer are struggling on designing UI, although some have come a long way, e.g. Apple, Samsung to name a few. But you would not believe the amount of research went into that over many yearsNot just technical research, but also human behavior research. I just wonder if car manufacturer have the capabilities to think this sort of stuff through from the drivers perceptive. Its part science, part art, part magic and gut feeling.

You said it there. UI designs are always evolutionary. But automotive infotainment is a platform which is trying to leverage the strides in mobile computing.

But again if you see, if you look at advances in driver assist systems (ADAS), that's a whole new ball game of UI. Here the end product is a UI overlaid on the windshield. So as you pointed out, its not yet another UI, but something which has to be very precise and intuitive. Because this UI would assume Mission-Critical importance. Processors providing compute power for such systems cannot just fail. However driving assist systems are a long way to go. But surely they are just waiting to happen.

Meanwhile for the state of the art infotainment systems, current UI infrastructure can work. Its just that the over all reliability of the mobile platform needs to be qualified for Automotive conditions.

[Jeroen]

Originally Posted by ampere

As I said in my earlier post, processors which power the cell phone can potentially find themselves on most of the low to mid end cars. But they have not been able to. Its not just the display alone.

The other reason is that the product life cycle of a phone/tablet is much smaller in time, as compared to a car. And once a processor is qualified for a car platform; that platform stays for a long time. Hence the revenue cycle time is much longer as compared to a mobile product.

Screen is also an added problem. Thats why if you see industrial displays are not very high resolution but much more error resilient. (Your own words.)

As regards to processor, it was never the performance but always the reliability. Still there is a basic performance requirement to power a decent display, run videos etc.

Because this UI would assume Mission-Critical importance. Processors providing compute power for such systems cannot just fail. However driving assist systems are a long way to go. But surely they are just waiting to happen.

Meanwhile for the state of the art infotainment systems, current UI infrastructure can work. Its just that the over all reliability of the mobile platform needs to be qualified for Automotive conditions.

I wouldn't look at mobile platforms, but as per my earlier post I would look at the technology employed for more then a decade in the aviation industry. Very, very high resolution screen with unbelievable reliability and robustness. The 15" screen of my little private plane, say a Cirrus or Diamond is better then my Mac book Pro and I can throw that plane around, land it extremely rough and it all still works, No problems on the screen no problems with the electronics. And what I would see in terms of dynamic data on a glass cockpit is a multitude of parameters compared to the most advanced car. Mission critical in aviation is a very different topic compared to mission critical for cars. The electronics coink out at 35.000 feet in the clouds is a very different situation from your dashboard going blank whilst doing 130 km/h on the highway. Also, planes system need to last for at least 10-15 years, so again, probably more stringent than cars.

I know a thing or two on certification of aviation systems, having been involved in some, and it's a very tedious, meticulous, check, double check, triple check and elaborate process. If it is good enough for aviation, General Aviation as well as Commercial aviation it is good enough for any other application, with the possible exception of space exploration, military and or under water deep sea exploration.

Have a look at this little package: http://en.wikipedia.org/wiki/Garmin_G1000

Just about every little 2-4 seater plane that has been built in the last decade is equipped with something like this. Far more elaborate than anything you need on a car. Far more reliable then anything you need on a car. Cost, depending on configuration upwards of a few thousand bucks. And a large part of that cost is due to the very specific elaborate certification that aviation instrumentation requires. And of course, the volumes are relatively low, certainly compared to cars, so its difficult to get a real scaling advantage.

Earlier this year I made a post here: http://www.team-bhp.com/forum/indian...g-cars-14.html

This is not science fiction, all this 'electronic' stuff is available and on offer today. Talk about mission critical! So even your average Ford is absolutely loaded to the gills with all sort of very elaborate sensors and endless computers and processors to facilitate all of this. So again, the processor is not an issue for the automative industry at all. Maybe the screens, to some extend, but that is probably more price then anything else.


So, I have great faith in what state of the art technology, both in processors and screens, can bring. I'm less convinced of the automative industry ability to make any sensible use of it. And that was really the point I was trying to make.

As to how this thread started I think a lot of it has to do with what drivers really want. What does your average Ford Fiesta want to see on his/her dashboard? Although i don't have any consumer insights I would expect it to differ hugely. My wife and I are probably a good example on how we differ in our expectations and what would make a good dashboard with good UI. Unless you can define within a reasonable bandwidth what your customers want and or like, it is near impossible to deliver anything that makes sense.

Jeroen

[Vid6639]

It's not just about the screens. Screens are relatively easier to change but screen is a cost part. Take a look at the BMW 5 series. The speedo and tacho needles are actually digital lcd screen and not real dials.

What is bigger aspect is an automotive grade processor that can handle the UI.

Audi has just tied up with Qualcomm for the first step to have LTE enabled connectivity in car:

http://www.qualcomm.com/media/releas...ectronics-show

http://www.redorbit.com/news/technol...reless-010813/

http://www.pocket-lint.com/news/1188...-car-broadband

The LTE chip is automotive grade which means additional cost and much more stringent tolerances. These are nowhere close to mobile grade.

Even for companies like Qualcomm or Apple to get into automotive is not easy because of the automotive spec for reliability. It adds huge cost and volumes are nowhere near as mobile market.

[ampere]

Originally Posted by Jeroen

I wouldn't look at mobile platforms, but as per my earlier post I would look at the technology employed for more then a decade in the aviation industry.

I meant only for the compute not for the display. As the infotainment requirements are similar to mobile.

Very, very high resolution screen with unbelievable reliability and robustness. The 15" screen of my little private plane, say a Cirrus or Diamond is better then my Mac book Pro and I can throw that plane around, land it extremely rough and it all still works, No problems on the screen no problems with the electronics. And what I would see in terms of dynamic data on a glass cockpit is a multitude of parameters compared to the most advanced car. Mission critical in aviation is a very different topic compared to mission critical for cars. The electronics coink out at 35.000 feet in the clouds is a very different situation from your dashboard going blank whilst doing 130 km/h on the highway. Also, planes system need to last for at least 10-15 years, so again, probably more stringent than cars.

+1 Totally Agree. But how much is the cost difference for these materials when it needs to go into low end cars. Can they afford the margins?

And a large part of that cost is due to the very specific elaborate certification that aviation instrumentation requires. And of course, the volumes are relatively low, certainly compared to cars, so its difficult to get a real scaling advantage.

I agree. All I am saying is along with the display, reliable compute for automative qualification plays its part. In the order of reliability :

Aviation > Automotive > Industrial > Consumer

So, I have great faith in what state of the art technology, both in processors and screens, can bring. I'm less convinced of the automative industry ability to make any sensible use of it. And that was really the point I was trying to make.

I do understand the point. My concern was not the cockpit part but the infotainment. What you say there makes sense. Infotainment is more like extension of our mobile world. Thats why now many cars have dual screens where one is mainly for infotainment only. (Please correct me if I am wrong).

As to how this thread started I think a lot of it has to do with what drivers really want. What does your average Ford Fiesta want to see on his/her dashboard?

My only view points were :

- Even car infotainment will need like what mobile systems needs
- Video/Audio Internet etc.
- Just like its stringent to get qualification for aviation, similarly in some capacity, it difficult to qualify the same mobile system for automotive.
- Totally agree to your point on availability of performance. Its the business model for automotive infotainment and its association to automotive qualified hardware which makes the difference I think.

EDIT:
- In context of this thread : I think its price margins for the screens. (In answer to OPs question on my Screens are so outdated)
- Would be good to get some details on this.
- As regards to playing content, the automotive class Apps processor requirements add up. Here I am referring to the overall infotainment system.

[Jeroen]

Originally Posted by Vid6639

It's not just about the screens. Screens are relatively easier to change but screen is a cost part. Take a look at the BMW 5 series. The speedo and tacho needles are actually digital lcd screen and not real dials.

Please take a look at the link I provided earlier about the Garmin G1000 Glass cockpit. the G1000 provides as it very most basic function a complete flight director. Providing one or two needles is childs play compared to that.

Originally Posted by Vid6639

Audi has just tied up with Qualcomm for the first step to have LTE enabled connectivity in car:

The LTE chip is automotive grade which means additional cost and much more stringent tolerances. These are nowhere close to mobile grade.

When I'm not busy on this or other forums or traveling I do a bit of work in the telecoms sector now and then. Could you please explain what you mean by a LTE chip being automotive grade? What standard is that? We have been known to get ourselves involved in LTE chip design. Our design team might have overlooked something?

Just for your reference you might want to look up how many cars today are already equipped with 2G and or 3G connectivity?

[Jeroen]

Maybe just a few examples to illustrate that pre-LTE technology (GSM, CDMA, WCDMA) is perfectly capable to work reliable on mission critical stuff in the automative industry.

A press release from 2009:
http://abcnews.go.com/Business/Autos/story?id=3706113

A company providing many such services on millions of cars, mostly through 2G, 3G and now slowly converting to 4G
https://www.onstar.com/web/fmv/home?g=1

Some more background on this company and some of their technology:
http://en.wikipedia.org/wiki/OnStar

You will note that they had similar services running even before 2G and are still running similar services on CDMA so, not really LTE.

Their main service is the ability to remotely shut down the engine. So pretty mission critical I would say, you don't want the engine to shut down at the wrong moment. So all of this was already available based on analogue technique in a time where we did not even know LTE?

Of course, LTE will provide many additional advantages over the previous technologies, but I'm not aware that there is such a thing as a LTA automative grade chip.

So, in essence we have literally millions of cars on the roads today equipped with various degree of connectivity for the purpose of multimedia and or direct interaction and or manipulation of the onboard engine management systems. All without LTE yet. So I'm a bit lost on this need for LTE automative grade chip.

In essence what LTE will do is provide a much larger bandwidth (i.e. considerable higher up and down load speeds). More relevant for mobile devices is that LTE chips in a properly tuned LTE Network are far more efficient, so less drain on your battery. Not a big concern for the automotive industry though I would think.

Jeroen

[Chethan B G]

I am with Ampere on this.

Quality grades as we all know ranges from (in layman's terms); Commercial grade - Industrial grade - Automotive grade - Medical grade - Military grade - Aeronautical grade - space grade.

We look at what is commonly available (Commercial grade electronics) and think that it is so simple / cheap, but is still not available in cars. But practically, it is not that simple - To design and manufacture similar devices with stringent tolerances, is not easy. That is one of the reason, duration between updates is longer in automobile industry.

But this trend is slowly changing. With recent cars providing BT, Touch screen video and some even a built-in GPS, commercial grade electronics is getting into / contaminating automobile industry.

The question is; Does that effect performance / reliability of the car in any way? It is only the entertainment part that is lower in quality - Is it not?

What happens when UI integrates all hardware in to the CAN-BUS? Will it still be simple?

[Vid6639]

Originally Posted by Jeroen

Please take a look at the link I provided earlier about the Garmin G1000 Glass cockpit. the G1000 provides as it very most basic function a complete flight director. Providing one or two needles is childs play compared to that.

Does this garmin unit have Bluetooth, iPod control, Facebook feed, restaurant finder and DVD playback?

If it doesn't, then it's not what the auto industry is looking for.

When I'm not busy on this or other forums or traveling I do a bit of work in the telecoms sector now and then. Could you please explain what you mean by a LTE chip being automotive grade? What standard is that? We have been known to get ourselves involved in LTE chip design. Our design team might have overlooked something?

Just for your reference you might want to look up how many cars today are already equipped with 2G and or 3G connectivity?

I don't need to cause I'm working in the same mobile field. I provided public links on the products I know of. There maybe more.

Regarding automotive grade in general following is done extra over normal mobile application:

1. Different packaging of IC conforming to auto grade
2. Different qualification with tighter spec
3. Different temperature range qualification

I will need to search on net for general details on this as what I know may not be publicly shareable.

When I'm not busy on forums, I spend some time in testing chips for various applications including ones that go into most mobiles today.