General Motors Company (GM) Earnings Call Transcript & Summary

All right. Great. Good afternoon, everybody, and thank you for joining us for our session with GM. I'm Itay Michaeli, Citi's U.S. auto analyst, and I'm delighted to have this session with GM to talk about Super Cruise and other automated driving topics. So from the company, we're very happy to have Mario Maiorana, Chief Engineer for Autonomous Development and Validation at GM; as well as Mike Szachta, a Director of Product Marketing and Connected Services. We'll keep this to a fireside chat session. Lots of interesting topics to go through. If you do have any questions for the team at GM, please e-mail me throughout the course of the session, and we'll try to get to your questions throughout. So Mario, Mike, thank you so much for being here.

Thanks for having us.

Thanks for having us.

Great. So let's get right into it. Super Cruise, obviously, a system GM has been developing for years. Maybe just -- and for those who are maybe a bit newer to learning about the system, maybe just review the system a little bit and particularly the defensing suite within it and maybe some of the differences in the system of what you just launched on the Escalade and some of the other Cadillac products, on the VIP electrical architecture platform versus what you'll be launching soon on the Bolt EV?

Okay. Yes, I'll take care of that. So it's always exciting to talk about Super Cruise. I got a passion for it, and it's great sharing this good news story with everyone. So Super Cruise is the industry's first true hands-free driver assist feature. I'll give them the sensing suite, but it's built on 2 foundational technologies: one of them being our high-definition LIDAR-generated maps, which are on board; and the second being our -- and we'll get into a little more detail on what that provides for us. And the second one being our, again, industry first driver attention system, which utilizes a driver-facing camera to determine driver attention and awareness. So those are the 2 foundational technology. We'll take a little walk around the vehicle, sort of a virtual walk around the vehicle to talk about the sensing suite. So on the front of the vehicle, in the front face, we have a -- somewhat centered as a forward-facing long-range radar, which is used very much for the vehicle detected ahead, for our longitudinal control, mainly for longitudinal control. And then we also have 2 short-range radars that are located on either side. So on the front, you got a long-range radar centered, 2 short-range radars on the side. Those short-range radars are used for object detection to the side, which we can use to dynamically offset from a vehicle, let's say, you're passing a semi, as well as they're used for the new feature, which we'll get into the lane change on demand. Also on the front of the vehicle, behind the windshield, is one of the main cameras, it's our front camera module. And that camera is used again for not only vehicle detection, but it's the main sensor used for line detection, marking the lines on the road. So that's another important sensor. On the rear of the vehicle, as we move backward, you have again a suite of radars in the back. You have either 3 or 4 short-range radars, depending on the geometry of vehicle. You have 2 that are angled out towards the side, again, use that dynamic offset or for lane change on demand. And then you'll have either 1 or 2 in the center, again, depending on the geometry of the vehicle, which also helps and is also key radars for some other active safety features like rear auto braking and such. Also located around the vehicle, you have 4 additional cameras. Many of you are familiar with these cameras. They're used for sort of that top-down 360 view you see. 2 are located in the side mirrors, one in the front grill and one in the back. We use those also as a redundant input for lane sensing. We can use those cameras to see where we're at in the lane, where the lines are, and those sensors are used mainly for redundancy. So then we have -- interior-wise, we have -- just making sure I cover everything. We have the driver attention system. That, as I mentioned, is the customer-facing camera. That works in concert with IR emitters that are in the steering wheel, along with that light bar that we use for HMI. And again, that is to detect driver attention, make sure they're head pose and eye gaze is forward on the road, because that is a key safety feature of Super Cruise, is ensuring driver attention and awareness. We need monitoring the situation, and we need to know that they're capable of taking all of our moment's notice if the system needs it. So some sensors that we also use that are less obvious is we do have a high-precision GPS to locate the vehicle on our high-definition map. That high-precision GPS is also corrected through a cellular network. That's a corrective GPS and it gets within 2-meter accuracy. So again, very high accuracy. So that -- and then obviously, we have the map data, which we already discussed. That map data is -- allows us to know the road we're on to make sure you're on a Super Cruise-enabled road as long as it gives us some of the road geometry. Whether the lane you're in is ending, the lane you're in is exiting, we need to know all this information so the system can operate properly. So you mentioned the difference between our first system and our VIP or the enhanced system. Essentially, the sensing suite is similar in its content. However, it's more powerful. We've improved the sensors. The sensors have better detection capabilities. As well as a big improvement is our compute platform. Our compute platform has a lot more bandwidth or throughput. So we're able to make a lot of enhancements to the system, not just feature enhancements such as lane change on demand, but also performance improvements so that the lateral control, longitudinal control are improved and feel -- what I'd like to say is a little more natural driving experience to the driver, better performance overall. So hopefully, that answers your questions. If anything else, we can dig in deeper.

That was very thorough. Maybe -- so thank you for the sensor suite. Now if you think about using them, it's our understanding that GM does a sensor fusion in-house. And I was hoping you could talk a little bit about your efforts and experience there and the decision to do in-house fusion as opposed to using suppliers. Or do you use some suppliers for some part of that stack?

Yes. That's a great question. I mean, the first big thing is this is a GM design system, right? We designed it internally. We determined the redundancies. We determined the sensor suite we needed. And we wrote a vast majority of the software for it. Now when we make a decision, right, we look at the specific area we're talking about, in this particular case, sensor fusion, as you referred to. And we saw a competitive advantage in doing it ourselves. We wanted the intellectual property in that space, and we wanted to really understand it. So when you're looking at lane markings or you're looking at vehicles ahead of you, it can be done with a single sensor. But safety being our pillar here, we wanted to take multiple sensors and use our in-house proprietary software to fuse all that information to paint the most accurate picture, I'll say, of our surrounding environment and the vehicles are there. And we extend that type of decision-making to other areas of the system, right? We'll make decisions on when we purchase a sensor or software for that sensor or when we decide we want to bring it in-house. Again, it has to do with if we feel like we have a competitive advantage in doing it ourselves. Much of the sensors are purchased as commodities. And we work with the supplier to communicate the information they're getting and bringing it into our software so that we can fuse it. Most -- all of our control, both lateral and longitudinal, is done in-house also.

Perfect. And then if you talk about LIDAR mapping, I think it's definitely been an approach that GM has talked about for a number of years. I'm curious around how you update those maps, and specifically, whether you are extracting any data from the onboard sensors to readily update the maps. I think about what Mobileye REM has done, for example, over the years. Is that something GM is doing with Super Cruise as well as many might do in the future? Just curious how you go about updating those maps.

Yes. Right now, the strategy for updating maps, and it's a good question, we do -- even in our initial system with the CT6, we did over-the-air updates of our map. We started doing them at a quarterly rate. We are bumping up to that. We're nearly monthly now in updates. So we do over-the-air update our maps. And there's good reason for it, right? You have road geometry changes through construction and so forth, and you need to go out and update those maps. At this point in time, it's all done through the LIDAR-generated map information. So we're not acquiring any data to update those maps at this point in time. We're always looking for improvements in the system. But right now, that's how we're managing it.

Perfect. And thinking about the ODD, the operational design domain, of Super Cruise. Right now, obviously, it's on certain highways where you've mapped. These are the [ widened ] roads. How far can the current Super Cruise system go in terms of the ODD? Can this generation -- we know GM has a goal to go beyond that. I'm sure what we'll get to that. But could this generation on the vehicles today, the last generation, go beyond compatible highways? And if not, kind of what percentage of compatible highway domains, these are over 200,000 miles, have you kind of [indiscernible] how far along are you on the compatible road?

Yes. I mean, we're always looking to expand our domain. We started, as I'm sure you're aware, with approximately 130,000 miles. And then in 2019, we added what I'll call divided highways with some limited crossings, which is another classification of road. We added nearly 70,000 miles. We're over 200,000 miles now. And we continue to look at other domains that we can operate in. We obviously need to balance that against the sensing suite and our control capability with safety being the #1 priority. We are looking to grow it. We will continue to grow it. And GM, as we've said, our ultimate goal is to have hands-free operation for 95% of the driver's operation.

Perfect. And from an OTA perspective, we talked about the LIDAR map update, but is this -- the new Super Cruise system, now on a more advanced architecture, going to enable sort of OTA updates on the, I guess, the ADAS controller to improve functionality? Are we there yet where we could see new perception software or even new sensor fusion-type software be OTA to the vehicles? Or is that -- is it mostly OTA or all the OTA really confined to the LIDAR maps?

Right now, in the initial system, it was all confined to the LIDAR maps. When we've gone into our VIP architecture, it does enable that, and we are looking at all of our options, like you said, whether it's fixes, enhancements, feature growth. Right now, the main focus is map, but that VIP architecture is really what's enabled us to start investigating all of our alternatives. Mike, I don't know if you had anything you wanted to add there because this is kind of in your wheelhouse a bit.

Yes. It's a great question. And this is -- as we're working through on the technology side, that's one of the values that we see as part of the offering to consumers, is that it's a platform that has a life cycle. The vehicle, from a future improvement perspective, can grow over time as we develop more and more capabilities. Not that we can really disclose those right now, but that's precisely the idea and the vision here, is to allow those that continue to be Super Cruise customers and users to expect more over time.

Perfect. And I know these are -- the newer version of Super Cruise is still relatively earlier days. But any feedback you can share from consumers, any early usage and performance data on highway driving that you've observed?

Yes. We have a lot of interesting data out there. I'm trying to make sure I have all the latest numbers. Right now, we have approximately 10 million miles of Super Cruise operation, obviously, mainly in the CT6 because that's where we have the most experience. We have found through our data that customers are utilizing Super Cruise about 50% of the time that it's available. Trying to remember the exact number. I think it's quite a few miles a week. I don't have that number right in front of me. But the interesting fact that we like to share with is there's been an overwhelming positive response from those customers who use it. Our surveys have shown that existing Super Cruise customers saying that the -- that Super Cruise -- 85% of them said Super Cruise will be a major contributor to their next purchase. And within that 85%, many of them said it would be a prerequisite to vehicle purchase. So those are the types of numbers you can't really ignore because you don't see 85% acceptance and many saying, it's going to have to have Super Cruise. I'm not even buying the vehicle, right? So we're seeing very positive response. And that's why we've made the announcement that we'll have it on 22 products -- 22 models by calendar year '23, because we've seen the customer acceptance and their joy in having the feature. So we want to get it in more vehicles. We want to expand it to more miles, which was a discussion you and I just had, as well as we got to keep getting them more features because their feedback has been heard.

Awesome. Maybe I'll add 2 more on the current system, and then we'll get into the really interesting go-to-market approach and even next-generation system. So first, obviously, I think Super Cruise is known for a lot of things, but one of them is the DMS, the driver monitoring system. And just hoping you can talk more about what you're doing there, how does it compare with the other approaches within DMS, and then ultimately, kind of what you see maybe in future generations of DMS usage going.

Yes. I mean, our driver attention system, or DMS, as you're referring to it as, it's a common theme. I've been saying, your safety was our #1 priority, right? We needed to ensure, being a level 2 hands-free system, which is what we wanted to be from the get-go, that we had driver attention in the system. And driver attention, we felt the best way to do it was with the system we've implemented, which, again, as I mentioned, is that rearward-facing camera, which is monitoring the driver while in Super Cruise to ensure eye gaze and [ head is ] on the road. We have infrared lights in the steering wheel, which help when it's dark. We can see through the vast majority of sunglasses. And then we've implemented a very -- what I think is a very intuitive escalation system. If you're not familiar, if you've been in a Super Cruise vehicle, right, we have the light bar on the steering wheel. And when the system is engaged and steering, you're a solid green. If you look away, tune radio for a little bit too long, that steering wheel will start flashing green. That's trying to draw your attention back to the road. If that does not work, we'll then go to a flashing red escalation and use either haptic or audible to get your attention to take the steering wheel. At that point in time, we're handing control back to you. We stay in control while your -- while the drivers getting their hands on the wheel and getting everything oriented. But we've asked them to take control back. If they try to engage it after that, we'll let it reengage, but they will know that the reason it disengaged was a lack of attentiveness. So we think that right now is -- and as I said, we've improved it from first gen to the current gen. I know in the first gen, we focused mainly on head pose. In this latest gen, we've added eye gaze as well as we have received some feedback of sunlight in that coming in, maybe hitting the camera and washing it out a little bit, resulting in inability to see driver's attention. We fixed that through software, some coat -- lens coatings, some shrouding on the sensor itself. So we're -- we've looked to continue to improve that. And where can we go in driver attention? That's interesting. There's a lot of places to go with it, and we really need to see because I personally see uses of it even within the nonautomated driving or maybe the active safety space. So I think we're really just starting to scratch the surface with driver attention, and hopefully, there'll be more to come and more to announce soon.

Yes. Just to add to that a little bit. The thing that -- in my experience in using the system, it's an incredible teacher of -- and I think about I have a young teenager who's going through drivers ed right now that if I could -- that's -- it's like having that parent in the next seat, make sure you're keeping your eyes, make sure you're standing around you. It requires that type of attention that it's -- we sometimes get complacent about. So I agree with everything Mario just said. It's -- there's a lot there that even in a different use case, that there's value in that feedback loop in making you a better driver. It's a different way to drive, for sure. And what I felt -- what I feel when I'm using it is I feel I'm in control and more aware of my surroundings because the system reinforces it so strongly.

Yes. That's a great point, Mike. I know from my personal experience driving it, I have a lot of miles in Super Cruise, obviously. It is a good training device. It seems odd to say you're even more alert when Super Cruise is engaged than when you're driving. But it seems to be true. And I think the reinforcement there is when I'm driving, I want Super Cruise engaged. When Super disengages because I looked away, it disappoints me, right? Because you enjoy the Super's operation. So to some extent, trains you to say, "Hey, I need to keep attention, or it's going to turn off this great feature." And then you don't want that. You want to maximize your Super Cruise model, especially on long trips.

Absolutely. I really appreciate the detail there. 2 quick questions we got from an investor. And I guess it relates to as vision improves, with the scope of -- would GM remove radar? Maybe talk about what will make an OEM want to maybe decontent sensors, let's say, radar in favor of vision? And then secondly, is there any vehicle-to-vehicle communication with Super Cruise?

Yes. At this point in time, there's no vehicle-to-vehicle communication. We'll answer that first. We're always looking to improve and always looking at new technologies. But right now, there's no vehicle-to-vehicle communication. The sensor question is an interesting one. I think as I said, we developed ours in-house. We looked at the sensing suite and decided what we wanted for redundancies and what -- built our safety case for the system. Obviously, as an OEM or GM looks at improving the system or changing the system, right, you're trying to balance the performance of the sensors, the domain you're trying to operate in as well as the cost of the system, right? Because if you can get the cost down, you can make it more available to more customers. So as you define your domain better, as sensors improve, right, that could drive the need for less sensors. If a technology of a sensor all of a sudden goes up significantly, right, then all of a sudden, you can depend more on that sensor. But -- so you've got really sensor performance, the domain you're operating in and then cost. You're all trying to balance that. And that's why it was important for us to do this design in-house, so we are comfortable with what we came up with.

Got it. And I think about the go-to-market for Super Cruise, maybe talk about the decision to offer Super Cruise as a subscription after, I think, the initial few year period. And kind of how did you arrive at sort of the right subscription terms and maybe the anticipated future take rates within that?

Yes. I think it's really an interesting question. And so when we've looked at the subscription, from the outset, when we brought Super Cruise to market with the CT6, we had a subscription component as part of the go-to-market. Because we knew that for the system to continue to operate safely and within the right parameters, we needed to have these OTA map updates and potential for software updates and new features to be added down the road. So it's been part of that decision. The right balance point is it's tough because you run into this situation where you want it accessible to as many users as possible, but there's -- it's such a new category that it needs -- there's time, right, that we felt that customers needed to come to the decision around, yes, this is something that I'm going to continue with or not. So we've tried to make it so that access to the system is for as long a period of time for people to make that informed choice. And then after that point in time, offering Super Cruise fits into this bigger idea of how we've been talking about connected services and all the products that we offer from OnStar to in-vehicle connectivity. So what we see now is that we give customers and empower customers for the choice at the end of that term to really get the most out of their connected vehicle. What we tend to believe and have seen to date is more and more customers that are in the segment so far want to -- want a fully contented connected vehicle. And that includes domains like safety and security from OnStar, connectivity that we have from in-vehicle or connected infotainment, et cetera, as well as automated driving coming from Super Cruise. So we see that with all the options available, more and more customers are going to be choosing fully contented.

Excellent. And on that point, we track attach rates a lot on our team. And it's interesting to see Super Cruise kind of move higher on the Escalade, and it's still early days. But as Super Cruise becomes a subscription, should we anticipate the attach rates to go substantially higher? Because you now have all this lifetime shot on goal, if you will, to earn revenue on Super Cruise. And actually, a second question we just got on e-mail is, with Tesla charging about $10,000 [indiscernible] a level 2 system, is GM leaving money on the table?

Well, I'll try to answer the second question first. We think that from a lifetime value extraction, I'm not going to comment on what other people are doing in terms of go-to-market. We feel like we're striking the right balance to reach as many customers as possible with what's included with the vehicle versus what we might charge for the service. And back to the first question, I'm sorry, I missed that one again.

Yes. Just as we think about the subscription opportunity, should we expect attach rates to go up? Because now you can make money a lifetime and you have more kind of shots on goal for revenue.

Right. Yes. So yes, it's a fair question. And I think from our perspective, the answer to that question is quite simply yes. As we -- as Mario talked about earlier, we're trying -- we're going to be -- we've already announced that we're going to 22 vehicles by calendar year '23. And that's really enabled by this go-to-market because it gives us a good deal of flexibility to maximize exposure into more and more vehicle models. Because we do believe when we think about it from the standpoint of where the category is growing, where we see that we're getting past the early adopter phase, right, so as you kind of go and grow, you get some efficiency from being able to drive down the cost curve and make it more approachable. So we do see that it's going to cause a great deal of additional pull-through in terms of contenting vehicles.

That's a great point. And then one more on just the go-to-market. So I think your subscription is interesting. I think right now, it's on a monthly basis, depending on what your OnStar plan is. Is there an ability to offer Super Cruise on a per use basis? Let's say, I'm driving, my map traffic data tells me I'm going to be stuck in some traffic jam for the next 2 hours. Is there an opportunity for like a onetime software unlock on an equipped vehicle? Just curious in terms of other kind of those types of opportunities.

Yes. So in terms -- we're always exploring all of those. Right now, we're -- we've established this capability, which is based on a monthly, but those can be purchased in annuals and things of that nature. We are always exploring, do we want longer durations? Do customers want this onetime fee? Or are they interested in shorter duration because it's situational? I think anyone who's ever dealt with the subscription business, it's a bit of a double-edged sword. You get a lot more appeal, but from an economic viability perspective, it's not necessarily the best approach. So we try to balance those factors appropriately because it puts a lot of burden actually on a customer to think through this idea of do I want to turn it on today? Or do I not want to? And just imagine if you tried to do that with your cell phone every time you wanted to use it. So we don't take that approach. We think that generally, customers that are accepting of the value proposition, which, as Mario indicated, is a very high portion of users that we've seen so far, that they're bought in. We don't see the occasional use as being kind of the thing that's a driver of any of those decisions.

Absolutely. So we talked about the current system and the go-to-market. I'd love to spend a few minutes on -- as much as you can share on the next generation. I know GM and I think we actually -- at our symposium last year, we talked with a company about Ultra Cruise and this path to go from highway to the company's goal of, I think, 95% hands-free driving -- or represent all driving scenarios being hands-free, which I imagine initially is an L2+. What can you talk about the evolution of Super Cruise into the next-generation Ultra Cruise system that GM is working on?

Yes. I mean, I can speak specifically to Super Cruise and some of the stuff we're doing there. But you are right in saying, right, our ultimate goal is 95% hands-free operation. But Super Cruise, we made a lot of improvements. As you know, the feature growth really hitting a steep part of the curve where we're adding feature, especially with our new VIP. In '21, we added our lane change on demand feature as well as some other things that were a little more subtle but important, like ease of engagement, and other feedback we got, the overall performance of the vehicle. I think it feels a lot smoother, a lot more natural. And now we're taking it to the next level. We've announced that the HUMMER would have auto lane change will be -- which will be system initiated as opposed to driver initiated. We're working to get Super Cruise working while towing. So there's a lot of features we're adding to continue to grow throughout the years, right? And another big one is, as we talked about, we expanded our roads from 130,000 to 200,000, and that's a big one for us. We need to stretch that domain. We need to work in more domains. But we need to roll it out in a logical manner, right? We need to make sure our sensing and our software are capable, that the system is safe and then we'll roll it out. The beauty of domain updates is often they're backwards compatible, I'll call it, right? We can update maps and just open new domains to existing customers. So that's really where I see it going is added features to provide more value, but even more importantly, is more domains. We have to open up the domains because that would get the usage up higher, I believe.

Absolutely. And maybe one other kind of, I guess, longer-term question. I think GM has kind of, I guess, publicly noted that there is potential for the Cruise team in San Francisco and the learnings there to potentially leverage that into future generations of sort of GM's Super Cruise and Ultra Cruise as you look to achieve those goals in the 95%. Just from your team, are you talking, working with the folks at Cruise? Is there a potential to port some of those learnings, whether it's sensing, mapping, driving policy. So just curious kind of whether there's a future synergy there between your teams.

Yes. Right now, the teams are separate teams, obviously, because of their technology, the sensing they use, the domains they operate in are dramatically different. With Super Cruise, right, we've restricted to domains of divided highways. We continue to look to grow that. So right now, the teams are operating individually, but stay tuned because there'll always be feature announcements as we try to better integrate and provide more and more features.

Great. Of all the questions, we actually received one. Maybe walk through, as you -- in the process of expanding the ODD and current Super Cruise, whether it's -- is it mapping and then testing and validation, kind of how long does it take roughly to add additional miles?

Yes. I mean, it is -- it does take time. Obviously, you have the mapping time itself, and that's highly dependent on what type of domains you're expanding to, right, how many miles you're adding. That's just kind of math there. But then internally, right, in some cases, you may have additional software development that's required because you may need to add some features or checks for that particular operating domain. So that, again, will vary the time. But definitely, there is extensive testing. Even without software changes, right, we're going to run those miles. We're going to run through safe cases where we get verification mileage under different conditions. Because, as you know, conditions widely vary, too, right, from a bright sunny day, which can pose its own challenges if the sun is forward-facing into the camera, to inclement weather and evening. So there's a lot of extensive testing done in any particular domain before we feel comfortable releasing it. So unfortunately, putting a set time frame, it's difficult because it depends how big the expansion is, whether software development is needed or whether we think we're confident operating in this domain, let's just test it, right? Then those are the easier ones. So it varies. But you've kind of hit all of the big points. You got your mapping time, you've got your calibration and tuning time or new feature development, if necessary. And then you've got your validation and verification, which is expensive because you need to cover all those edge cases with different environments and so forth.

Absolutely. That's super helpful. And then one kind of safety question because you referenced it a number of times. And I guess with Level 2, Level 2+, there isn't, I guess, a formal kind of rule for what safety needs to be in terms of metrics. How does GM kind of measure that for Level 2+? And today, if somebody does disengage Super Cruise, are you kind of getting feedback on the type of information?

Yes. We do get some information on state changes and engagement and how Super Cruise is being used and where it's being used. But from a safety perspective way, I said, our key always goes back to the driver attention system, right? The key of any Level 2 system is driver attention, driver engagement. They are in control, right? It's a driving assist feature at that point in time in Level 2. So the key for us is that driver attention. We need to make sure the driver's engaged and their eyes are forward. We're going to try and provide hands-free as much as possible as long as the driver's attentive. We should be able to provide it entirely within the domain if they're in the right domain. Now there are circumstances which makes driver attention important. For example, you're driving an evening and you run across a snow squall if you're living in the north, right? And all of a sudden, the lane lines are obscured. Well, I always tell people, if you can't see it, then the system is going to have a hard time seeing it, right? And that kind of stuff can happen in a moment's notice, right? So our safety premise has always been, we need that driver. It's a Level 2 system, and we need to make the right decisions at the right time to pull them in and have them engage and take over the system when the time is appropriate. It could just be a lane ending, right? You don't -- you're not familiar with the road. You don't realize your lane is ending or exiting. We need to bring you in. And we do right now about 0.5 kilometer before, figuring that gives you time to move over. But then there's the unpredictable ones like weather and other drivers on the road, right? So we need to be able to pull the driver in. So that's really the foundational piece for our safety is driver attention because it is a Level 2 system.

Absolutely. I think 2 last ones in the last minute we have here. What can you roughly share or talk about the cost of Super Cruise? And second, more at an industry level, we see a lot of new sensors, radars, LIDARS, higher megapixel cameras. What's sort of exciting to you at an industry level, away from maybe even talk about future product in the next few years in terms of what's happening, in a sense, or what's its capability, costs or both?

Well, I think cost is a big one. We're working hard to make this system more and more affordable. A great example of that is when we started, we were on Cadillacs. And now as we've announced, we're on a Chevy Bolt, right? So we're clearly able to put it on some of our lower-trim levels, right? We've taken the system cost from $5,000 option in the CT6, down to depending on the brand between $2,200 and $2,500. So that's a significant decrease in cost to the customer. So while that's not technical, I find that exciting because I get the feature into more hands, right? Just simply put, when I put it on '22 models before calendar year '23, it's going to be out there and to become the norm, to some extent, on all these vehicles, right? So that's exciting. On the technology side, yes, you're always looking for sense of improvements, but I think a lot of the rapid growth right now is within our software and controls area. We're adding a lot of features even with the existing sensor set and its capabilities. But the amount of features that could be added that you can kind of dream up, right, and create the software, that to me is the most rapidly growing area right now that excites me, right, is getting it out to more customers, expanding that road and adding features through software. Very, very fast-growing space.

Terrific. I think we're just about out of time. Mario, Michael, thank you so much. We've learned a lot. Really appreciate your participation, great conversation around Super Cruise, some exciting stuff out there. So I would thank you again for participating.

Okay. Thank you.

Thank you. Great questions.

Absolutely. Thanks, everybody. With that, we can go and conclude the session. I want to thank GM again for participating in a symposium, and have a great day, everybody.