00:00:04: Welcome to the Human and Technology Podcast.

00:00:07: This podcast is for anyone who develops, distributes or uses technology For all those that always have a feeling that technology overwhelms or dominates them... ...for everyone wants know how to deal with technology in everyday life.. ..For Anyone Who want's understand what technology does to us And How we can get our lives back!

00:00:33: is for those who want to make technology sexy.

00:00:37: All the product developers, designers, UX-UI professionals, product managers, CTOs and CEOs... And it's for you!

00:00:47: My name is Dr Peter Oreska.

00:00:50: my friends call me Dr Peter.

00:00:53: I am your host ...and i'm happy that are here.

00:01:00: Hello hello and welcome to The new episode of THE HUMAN technology podcast.

00:01:10: At the moment I am developing a new keynote and it will be on the biggest challenges we have in the automotive HMI world that i see, probably our will be giving it uh in the mid of june at The WeConnect Car HMI event.

00:01:37: But since you are my dedicated podcasting audience, My listeners You will get a little spoiler up front.

00:01:45: So I'll talk about basically the content that i'm thinking About.

00:01:51: It's all under development.

00:01:53: Nothing is fixed yet but at The end of the day and I Will Talk about Things That Come Up in Automotive HMI Development Challenges that I see and challenged us that might not yet have been addressed properly.

00:02:15: Designing your machine interfaces in the automotive domain today, I mean has never really been trivial but it is absolutely changing challenging.

00:02:30: for four decades development followed a relatively clear pattern more functions, more technology.

00:02:36: More possibilities.

00:02:38: ISO ninety two forty one hundred ten for usability as a basis the two hundred and ten for user experience.

00:02:46: yes all this gave us pretty clear path that we should go way to go.

00:02:56: The HMI served a analog digital, digital analog translator in the car.

00:03:04: So humans we are analogue beings doing analogue things and having analogue fingers and voices.

00:03:12: then this piece of glass on the microphone touch on the glass speak into the microphone.

00:03:19: than the HMI turns these into digital data that is delivered to black box, a digital blackbox and something is happening there.

00:03:33: And the result of that what is happening here is brought back to this piece of glass or the loudspeaker.

00:03:42: so then we can see something on our screen or hear some kind of beep tone or voice or speech.

00:03:50: all those things are traditional The this link function that the HMIS was sometimes better, sometimes worth but basically... ...the core role was stable.

00:04:11: These times may be over soon!

00:04:16: We are experiencing a Simultaneity of disruptions that has never existed in this form and our industry before.

00:04:28: Automated driving up to autonomous driving is fundamentally changing the role of humans on vehicle.

00:04:35: Our official intelligence questioning entire interaction paradigm.

00:04:41: New players, especially from China are bringing completely different ideas of what a vehicle HMI should look like.

00:04:50: What the vehicle really is and how a vehicle actually should work?

00:04:56: And that is only the tip of the iceberg.

00:05:00: beneath this surface further challenges are emerging That our at least as profound From software defined vehicles to the questions of the right level off display and interaction two connected HMI ecosystems that are existing.

00:05:19: And have a long since since long extended, yeah the HMI is going beyond the vehicle itself what as often overlook.

00:05:33: it's not longer just about usability trust, it's about control.

00:05:39: It is about cognitive relief and ultimately the question of what role humans still play in an increasingly autonomous intelligent distributed system?

00:05:58: In this episode I provide overviews on what i consider greatest challenges in automotive design where I will explore individual aspects in more detail, or dive deeper into details.

00:06:17: and yeah so this one is just like an overview episode.

00:06:25: One thing's clear if we do not truly understand these challenges We'll continue to build technology that can do everything except fit the people who are supposed use it.

00:06:39: yet today This introduction episode is on the greatest challenges and automotive HMIs.

00:06:47: The starting point of a cool series where I will elaborate the meta themes.

00:06:54: today, i have three points that i'd like to talk about.

00:07:00: number one are the three obvious challenges we have.

00:07:05: then two the five underestimated challenges the one biggest greatest outstanding challenge that we have.

00:07:16: All right, three obvious challenges five undamaged challenges and The One META Challenge.

00:07:28: Let's start with the obvious challenges automation artificial intelligence and HMOS from China.

00:07:37: Automation when The system, the car and technology suddenly drives.

00:07:47: Automation in cars is nothing new.

00:07:50: until now it has primarily taken over secondary tasks from automatic gearboxes rain sensors light sensors electric window lifters.

00:08:03: Now the character of automation is fundamentally changing.

00:08:10: The system intervenes into the actual driving task.

00:08:18: So at a moment, we are changing the car relationship between car and driver but when they're rolling robot on the human brain so what?

00:08:29: We had changing this correlation ship And then this means that vehicle shifts from a tool or passive to to an actor, and active actor.

00:08:43: So we move from interface... ...to interaction on almost equals And this massively changes the role of the HMY.

00:08:55: It's no longer just about usability bad about user experience and use a light but it is about central questions Who is in control at that moment?

00:09:06: who is driving?

00:09:09: Yeah, what is the system doing and why?

00:09:13: Is this system doing this?

00:09:16: when does the human need to intervene?

00:09:18: so basically What will happen next?

00:09:22: next in a near future in the next couple of seconds maybe up to a minute.

00:09:27: But well happened these are things we definitely need to display To communicate When We talk about HMIs in automated or autonomous vehicles.

00:09:44: so the HMI becomes more a mediator between humans and machine.

00:09:49: And then critical points are communication of states, handover from driver to car which I see a little less critical but definitely critical is the handover from.

00:10:09: The car to the driver.

00:10:11: so what would I drive with my level two automation and on German Autobahn?

00:10:17: And suddenly, the lane markers and then the system pushes these steering task back to me.

00:10:26: sometimes this comes as surprise.

00:10:31: we need to communicate.

00:10:34: How will this relationship with current driver change in the near future?

00:10:43: Headovers, as I described it are not an exception.

00:10:47: They're the norm.

00:10:48: they happen frequently and ambiguity means uncertainty.

00:10:52: so It is a safety thing that we were talking about here.

00:10:58: Automation automated car all cars only work if users understand what's going on at least in terms of its behavior, we will never really understand what an automation algorithm is doing.

00:11:13: What's happening on the software level?

00:11:15: but... The behaviour!

00:11:17: We need to understand the behaviour and be able to predict it.

00:11:22: so.. ...we need create transparency instead of black boxes.

00:11:29: We need make system-behaviour anticipable Is this word existing?

00:11:40: Able to anticipate.

00:11:41: We need the systems able to be anticipated, we need a context-sensitive constant continuous feedback.

00:11:56: so with automation automotive HMIs are fundamentally changing from an interface system that organizes control That organizes responsibility and that creates trust.

00:12:13: So we as HMI designers, We are no longer just designing interfaces... ...we're designing the collaboration between humans and machines And I've always been critical about saying okay.. ..we have a collaboration with human and machine because The human is definitely superior.

00:12:34: being needs to be master and machine is the slave.

00:12:40: With AI coming in, I'm ready to use the word collaboration because yeah.

00:12:46: The machine makes suggestions and improves and so on.

00:12:49: So already talking about the AI although my theoretic is still at the automation where you can see it's all somehow connected.

00:12:58: It sits on that not really independent.

00:13:02: number two of the three big challenges artificial intelligence AI from interface to an interaction partner.

00:13:17: Artificial intelligence is entering the vehicle on a rapid pace, voice assistant personalized functions learning systems that there are no longer futuristic concepts.

00:13:31: they're at least visible if not already implemented.

00:13:35: The real disruption lies deeper.

00:13:40: AI does not just change its function.

00:13:43: it changes the principle of the interaction itself, from a deterministic system to probabilistic systems.

00:13:55: We are not talking about one-to-one input output kind of thing anymore but that changes itself and interprets it seems to know.

00:14:07: So put in different way traditional HMI's follow clear rules you make an input then get a defined response.

00:14:16: AI systems work differently.

00:14:18: They interpret, they evaluate and anticipate And results are no longer always clearly predictable For users.

00:14:27: this means there is less direct control There's more interpretation by the system.

00:14:38: This is critical uncertainty in behavior of a system.

00:14:44: So we move into from menu-driven interactions to relationships.

00:14:54: With AI, the interaction shifts from traditional menus to dialogues for operating or collaborating and being a shared tool to be an apparent partner.

00:15:12: The HMI becomes Interaction Partner with all of these consequences Expectation management, biggest challenge for me.

00:15:23: We need to manage the expectations of the user and I see this as extremely difficult.

00:15:31: what can the AI really do?

00:15:33: Yeah we need to communicate this but other limits.

00:15:36: where does it work reliably?

00:15:40: or why not?

00:15:41: so expectations that are far too highly to frustration leads to wasted potential of AI, HMI's and maybe even safety critical features.

00:15:58: So we need transparency instead of magic here not what AI systems look like today as a black box.

00:16:08: this Black Box paradigm in a vehicle may turn out to be problematic.

00:16:13: the HMI must make its decisions Understandable, needs to communicate them and make them understandable.

00:16:22: Needs to communicate uncertainty.

00:16:26: it needs to keep the limits of itself visible.

00:16:35: Don't get me wrong not everything needs to be explained but enough needs to explain to enable trust.

00:16:46: To summarize, AI elevates automotive HMIs to a new level.

00:16:51: from a pure interface... ...to system that interprets, decides and actively interacts.

00:17:00: The central challenge is not the technology itself but how do you design an interaction that feels intelligent without taking control away humans.

00:17:23: The third obvious biggest challenge and other modifier that I see, China HMOs.

00:17:32: when different worlds collide Chinese vehicles are entering international markets at high speed And they bring their own HMI concepts with them.

00:17:45: the problem is these concepts simply designed in a different way.

00:17:54: they are based on different usage reality.

00:17:58: Different user experiences, different use cases, different contexts.

00:18:03: so things that differ.

00:18:04: there we have different users and the user group... In China their age of people who buy first car is something around thirty five or something, the average buyer.

00:18:17: Every car buyer age?

00:18:18: No not that first car.

00:18:19: in general The Car Buyer Age is in the mid-thirties.

00:18:23: In Europe US it's in the Mid fifties.

00:18:26: That there was one reason but... ...in China.. ..the car Is part of the digital ecosystem people have.

00:18:38: So you have your smartphone You have a tablet and smartwatch And then you've got your car which also needs to be part of this ecosystem.

00:18:51: This means in China, it's digital first with a high feature affinity.

00:18:58: features are always welcome.

00:19:02: the strong integration services entertainment building-of-ecosystem In China our different expectation regarding personalization speed feature scope and This then leads to HMIs that are from a European, US view significantly more feature-rich.

00:19:26: More too much sometimes playful and definitely more software centric.

00:19:34: for many European or U.S uses this quickly appears overloaded unclear simply irrelevant.

00:19:44: the real problem is of translation.

00:19:50: The transfer of these HMIs into Western markets is often not well done, let's put it that way.

00:20:00: It starts with the sheer language.

00:20:03: Yeah?

00:20:03: It's clumsy and incorrectly localized.

00:20:07: A few months ago I was sitting in a couple different Chinese cars... ...I'm NOT a native speaker.

00:20:17: They were all in English and then I'm not a native speaker, but even I felt that this language is definitely not correct.

00:20:24: It's not fluent it's nice its not elegant.

00:20:28: That's where it starts.

00:20:30: It continues with the visual design.

00:20:32: Yeah, it's not aligned with Western expectations.

00:20:35: color sets icons The good general overall quality had the interaction logic as a problem.

00:20:44: Somehow it seems to be a little unfamiliar sometimes counterintuitive.

00:20:50: Yeah, and then there are features We don't understand.

00:20:55: I just didn't understand what is this for?

00:20:58: And maybe This makes sense.

00:21:00: probably does make sense in China with the bigger ecosystem.

00:21:03: but yeah other features of their use cases or that we talked about difference in China But they won't work in Europe.

00:21:13: At the end, we have a system that is technically usually absolutely impressive but not really easy to understand and accept.

00:21:30: There's strategic conflict.

00:21:33: many Chinese OEMs face a dilemma The domestic market, the Chinese market.

00:21:41: The China Market is the main driver.

00:21:43: this where they sell a lot of cars.

00:21:44: This Is Where They Want To Have A Footprint?

00:21:48: The International Markets Are Not Really In The Car Focus That They Want to Go To Europe and They want to go to US However Successful There May Be.

00:21:58: But These Markets Still Need To be Developed And in Practice This Often Means We in Europe, we get an adapted version of a HMI which was originally primarily developed for China and only adapted on the very low levels.

00:22:19: And this is exactly where problems lie.

00:22:23: The challenge not to just translate Chinese HMIs... ...the challenge is re-think them based on the respective users markets cultural contexts use cases traffic concepts driving behaviors and so on.

00:22:44: this is a good message for people like me.

00:22:49: We can do that job.

00:22:51: we can help Chinese OEMs and Chinese first year suppliers to understand the European market into adapt their solutions so that they fit nicely into our world here.

00:23:12: A globally successful car maker, a globally successful automotive HMI is not just created by exporting something but by developing an original fully adapted intercultural design for the different markets.

00:23:38: All right, so those were the three biggest challenges that I see at the moment.

00:23:45: let's move on to the five challenges That i feel are very often not in The focused but they also have a strong influence.

00:23:58: and number one our software defined vehicles.

00:24:03: the HMI becomes a moving target.

00:24:06: so vehicles are evolving into software centric system and they continuously change.

00:24:13: I always compare this with the situation we have in smartphone market, if you all turn off our smartphones and put them display upside by side it is very hard for us to say which one is which brand or operating systems sometimes cannot even detect your own Smart phone between all the cut there.

00:24:33: absolutely similar of course.

00:24:35: they are different display sizes and nano frames.

00:24:37: And then whatever.

00:24:38: at the very end of the day, They look very similar and The difference comes with software.

00:24:48: so which apps are you having?

00:24:50: Which color scheme do you apply?

00:24:51: how does your start screen know what?

00:24:53: which widgets Do we have one on the main screen?

00:24:56: Yeah, all these things they make the difference and cars will follow.

00:25:00: The same way so that they look more or less the same in our software driving software platforms Or hardware platform for software.

00:25:10: And then the user experience the branding.

00:25:13: everything comes through Software And you can adapt it.

00:25:18: You could change and how often do you find, oh there is a new icon on an app or they have changed the HMI of an app?

00:25:27: This will also happen in cars.

00:25:30: so interfaces are no longer static but regularly updated.

00:25:37: So users must not only orient themselves once But continuously over lifetime And consistency becomes a central design task across the different HMI versions.

00:25:52: I can foresee that we will have a couple of HMI paradigms, they are fixed for a single HMI and then on top of that there'll be changes in adaptions.

00:26:02: so it's a delicate balance between the consistency.

00:26:07: Things are the way I expect them to be things.

00:26:10: Are they off always been?

00:26:11: Things are here where there somewhere else.

00:26:14: now, this is consistency and This becomes a central design task too.

00:26:22: balance this with With the flexibility that you have in software.

00:26:27: so The challenge at the end is not innovation, not only innovation but stability in change.

00:26:38: This is what I see as critical.

00:26:43: number two peak display.

00:26:45: yeah less suddenly more this.

00:26:49: a few years ago we had these A to A pillars displays.

00:26:54: the complete dashboard is a display and that there has where you do not have this.

00:27:05: Today it looks totally different, there are a few constant habits but most of them having more or less traditional architecture with the display behind steering wheel another one in the center stack on top of the centerstack and maybe head up display.

00:27:20: so we've reduced the display.

00:27:23: real estate vehicles Probably because we found out all digital is not good.

00:27:32: We're analog beings and most of the surface does not automatically mean better usability.

00:27:38: at The end, we had a problem that... ...we have to fill space there.

00:27:45: There was a lot of empty space on many of these display concepts Because they just didn't know what to put there Which is exactly opposite of what we had twenty five years ago where a six-inch display was the big one, that it wasn't.

00:28:00: A luxury display of six inches.

00:28:03: and we had to decide what do we show?

00:28:07: which way?

00:28:10: because there is limited space?

00:28:13: We are returning this.

00:28:18: It's good exercise or positive exercise about reduction in prioritization That becomes crucial again.

00:28:27: We must design a way that important formation is easy to grasp and quickly understand, then we see this tendency in the revenge of analog physical controls regain importance.

00:28:45: so focus shifts from show everything.

00:28:52: Number three, multi-modality.

00:29:00: We are having touch voice haptic gestures.

00:29:05: Yeah technically it is all possible.

00:29:08: the real challenge lies in The meaningful combination not just adding another channel Not just putting something on top that already works.

00:29:19: So so there's a clear assignment Is which Channel this for?

00:29:23: Which task?

00:29:25: where does It create a Real Value?

00:29:29: We need to avoid conflicting interactions.

00:29:34: At the end, it is about the orchestration instead of adding input.

00:29:43: so multi-modality only works as an fully integrated orchestrated system not a meaningless add on or features.

00:30:00: Number four, trust and transparency.

00:30:05: As system complexity continuously increases the trust becomes a critical success factor.

00:30:15: so systems states must always be clearly understandable.

00:30:19: visible decisions must remain traceable.

00:30:24: uncertainty must be actively addressed Because at the end, without transparency.

00:30:32: there is no acceptance.

00:30:33: There's only rejection and limited transparency is a safety issue.

00:30:45: And number five of them from my point-of view unnarrated HMI challenges are ECONOM systems.

00:30:52: The interface leaves the car.

00:30:54: we used to have one two One device in one location with one function and that is gone.

00:31:03: So HMIs are leaking into everywhere, so the vehicle has not a... The only location anymore of an interaction with the vehicle.

00:31:15: We can interact it from our apps or PCs And maybe someday we have screen on refrigerator, and then we can check the car whether it's charged or not.

00:31:27: Or if there are any software updates that we could install... We don't even need to get into this car for do-this.

00:31:34: This is an HMR ecosystem as I call them liquid HMRs.

00:31:39: Problems are that you have to create a consistency across different interaction instances like the vehicle Like the refrigerator screen.

00:31:54: We need to have a context dependent and role-based use in this one.

00:31:58: Where are we?

00:31:59: Are we, let's say safe situation at home on our sofas that we can for example think about different design themes and personalization and download them then into the car once you're in the car?

00:32:12: there are different things important should be done here so I see shift of complex interactions that are not necessarily while driving to external devices.

00:32:28: The HMI becomes, at the very end a distributed system and not just single one instance interface.

00:32:41: These five challenges I've mentioned are less visible.

00:32:45: They're strategically decisive because they do NOT concern individual features but the fundamental architecture of modern HMIs.

00:32:59: HMis become dynamic, reduced connected trustworthy and designed as a system.

00:33:09: so the future of automotive HMIS is not decided by major innovations but by the ability to master these sometimes subtle but definitely structural changes.

00:33:30: All right, I talked about the three obvious challenges and the five not so obvious challenges in automotive edge of my design.

00:33:39: And now it's not talk about them made at challenge Not losing The human in the systems.

00:33:47: with all these technological developments one fundamental question arises Who are we actually designing for automation?

00:33:56: AI soft verify vehicles more and more features.

00:34:01: All of this creates enormous momentum.

00:34:05: But, This is exactly where the greatest danger lies that we build systems That can do more and More but fit humans.

00:34:15: human needs humans ideas Human capabilities less and Less.

00:34:21: And then central challenge Is not a technological Challenge It's a conceptual Charge.

00:34:27: it's Not The question.

00:34:28: what is possible?

00:34:32: creates value.

00:34:34: A good automotive HMI is not defined by the number of its functions, but by that value it delivers to its users and what does value?

00:34:48: Technology in a vehicle must concretely improve human lives –not abstract-but in the context of use.

00:34:56: for example faster completing tasks faster more efficiently making them safer Reduce distraction, make critical situations more manageable.

00:35:09: Creating focus.

00:35:11: Highlight what really matters and hide does not matter.

00:35:15: Making life simpler.

00:35:17: Reducing complexity instead of increasing it.

00:35:21: And everything that doesn't serve to these goals is just a ballast.

00:35:31: It's useless cargo.

00:35:34: Many many current HMI's do not suffer from too little but form to much Too, many functions too.

00:35:41: Many options two mini interaction pass.

00:35:44: the result is not added value But cognitive load up To cognitive overload.

00:35:54: The human as the benchmark Not technology that did.

00:35:59: the real meta challenge?

00:36:00: Is is to cons consistently placed humans at the center.

00:36:06: Capabilities, limitations, expectations users contest dreams wishes all these are very human things and they must be the drivers of development.

00:36:16: And technology must be aligned with this not the other way around.

00:36:22: The future automotive HMI will not decided by next technology but right priority value instead features best technologies is not one that can do them most but the one that makes people's lives noticeable better.

00:36:45: All right, let us come to an end of this episode.

00:36:48: The Automotive HMI is at a turning point.

00:36:50: We are seeing more technology and possibilities And more innovation than ever.

00:36:55: That is exactly why one thing has become increasingly clear With previous ways we will not get any further.

00:37:04: If you want shape This development properly It's not enough to optimize existing concept.

00:37:11: We must fundamentally question these existing concepts away from feature thinking, away form isolated interfaces, away from technology for the sake of technology toward a system that truly understands human towards interactions that build trust towards solutions That deliver real value.

00:37:36: if you want to get this right we need to rethink HMIs fundamentally.

00:37:43: This is what this episode was about, and I will drive in the next weeks deeper into a couple of these challenges that are just mentioned.

00:37:54: Sharpen perspectives discuss concrete approaches on how they design this?

00:38:00: And yeah, how do handle automotive HMis in an increasingly complex world?

00:38:10: so... Final sentence.

00:38:13: The future of ability is not decided on the road.

00:38:17: It is the interface between humans and technologies.

00:38:23: If you liked this podcast episode a new thing, it's worthwhile for others.

00:38:27: Why not recommending it to a couple of your peers?

00:38:30: Your friends or colleagues?

00:38:32: The more people who listen to these podcasts the higher I will be ranked And even More People Will Be Able To Get My Thoughts.

00:38:46: That's it for today.

00:38:49: Thank you for spending time with me.

00:38:51: I hope to are able to take something with you and do something for yourself.

00:38:56: that will be forever.

00:38:58: For an unknown exchange, we will find me on a LinkedIn And my websites beta minus rusker comm and beyond minus HMI dot de.

00:39:11: write me an email under podcast at Beyond Minus HMI DTE.

00:39:17: Do it next time, take care and stay healthy.