top of page
Search
  • Writer's pictureMike Entner

You Can't Make an Omelette Without Breaking a Few Eggs: The Software Defined Vehicle (SDV) Paradox

Updated: Dec 3, 2023

By Michael Entner-Gómez | Digital Transformation Officer | Entner Consulting Group, LLC.



Picture this scenario: I’m driving my truck when suddenly a sensor in the emission system malfunctions. The engine stutters and the vehicle grinds to an abrupt halt. The car behind, unable to react in time, crashes into my rear. This isn’t a scene from a futuristic novel; it's real-life, and it happened. Now, let me pose a pointed question: How did all the vaunted safety certifications, which the industry heavily relies on, fail to prevent this accident? The answer is unsettlingly simple. Our automotive technology, despite being laden with decades of engineering expertise, is struggling to keep pace with modern needs. In this situation, it actually caused the accident for a non-critical subsystem. Why is this the case? The answer lies in the fact that the systems in a vehicle are predominantly designed by mechanical and electrical engineers. These engineers, undoubtedly brilliant in their domains, are linear thinkers by nature and not necessarily trained to create holistic architectural frameworks. They are deeply entrenched in the automotive industry, often with limited exposure to cross-industry innovations, leading to a sort of tunnel vision. After pouring over countless whitepapers, designs, and architectures, I've yet to encounter the metaphorical 'broken egg' that signifies true innovation. Most concepts are either recycled or iterative, failing to tackle the fundamental issues that led to my hypothetical accident.


Competing Worldviews


The core issue here, and let's be clear about this, stems from the starkly different worldviews of electro-mechanical engineers and software engineers — the latter being my own tribe. Traditional automotive engineers, steeped in their ways, really ought to take a step back and give software engineers the space to shine. To drive my point home, let’s pivot to the academic sphere for a moment. Imagine you're itching to dive into an automotive engineering or design program online. Your search will likely be a frustrating one. Such programs are as rare as hen's teeth. Now, flip the script and consider the abundance of online software engineering programs, right up to the Ph.D. level. Ever wonder why?


Here’s the thing: software engineers swim in the sea of abstractions, not tethered to specific, tangible applications. This contrasts sharply with the automotive engineer’s fixation on tangible, hands-on experiments — their benches cluttered with breadboards, wires, and actuators. But what’s the big deal with building to purpose, you ask? In theory, nothing. In practice, however, it's a one-way ticket to a labyrinth of 'standards' necessary for system interoperability, leading to a morass of siloed technologies and software hopelessly intertwined with hardware. This is the quagmire that throttles innovation, trapping it in the quicksand of over-specification and narrow vision.


The Boys (and Girls) Club of Automotive


Let's face it, there's a prevailing notion in the automotive industry that it's some sort of closed club – a belief that outsiders just can't grasp the intricacies of this world. I beg to differ. It's not about where you come from; it's about how you think and your willingness to learn. Take any seasoned consultant, for example. They could jump from crafting the perfect PB&J to engineering interstellar rockets, and still manage to pinpoint the crux of any issue. They dive into the deep end, unravel the business goals, spot the gaps, and chart a path forward. Sure, they'll need to get their heads around a few basics, decode the jargon, and grasp the essence of a design versus its execution. But give them a minute, and they'll be steering the ship with confidence.


That's me in a nutshell. I'm not your typical automotive industry insider. I didn't waltz into my bachelor’s degree until my 30s. But what I do have is a lifetime of hands-on experience – tinkering with (and eventually rebuilding) bikes, cars, and motorcycles since I was 16, out of sheer necessity; a.k.a. poverty. Now, add to that advanced education and experience in information technology, software engineering, and knowledge management and my desire to focus on SDV clearly emerges. It's this unique blend of practical know-how, from metalworking and circuit board crafting to parts design, combined with the finesse of manipulating data, that allows me to view the world through a different lens. It's about merging the tactile world of physical creation with the fluid dynamics of software, and that's where the magic happens.


Beyond the Body Panels


Alright, let's shift gears a bit here. You might be thinking, based on my previous musings, that simply bringing in a bunch of software engineers is the silver bullet for the automotive industry. Hold your horses, though, because that's not quite the full picture. Sure, software engineers are wizards at getting the bits and bytes in line, but they're not the puppeteers of data beyond the vehicle's confines. Alongside the term 'Software Defined Vehicle' (SDV), you'll often hear 'connected vehicle.' Now, this is a game-changer. It separates the automotive world from, say, a Mars mission, which, despite its complexity, has a singular purpose and operates in a relatively controlled environment. What's more daunting – the void of space or navigating the chaos of downtown San Francisco?


But let's circle back. Cars, especially in this 'connected' or rather interwoven reality, demand a whole new playbook. They morph into nodes in an intricate mesh of systems, all interacting at breakneck speed and scale. This is where networking and telecommunications come into play, and herein lies another gap for the automotive sector. Grill a seasoned automotive engineer about cellular tech in cars, and you'll typically hit two walls – telemetry and over-the-air (OTA) updates. If they dabble in infotainment, they might throw in a few extras like Bluetooth, streaming media, and the like. On an exceptionally good day, you might even stumble into a conversation about Vehicle-to-Everything (V2X) communications. But that's usually where the road ends. What we're seeing here is another case of specificity clashing with abstraction. The telecom component is often viewed merely as a tool, a means to an end, rather than as a gateway to a wider array of capabilities or as the backbone of an intelligent automotive network.


Alright, let's take this thought train a bit further down the tracks. Moving beyond the idea of merely connecting vehicles for software updates, fleet tracking, or collision avoidance, we plunge into the vast ocean of data aggregation, analysis, and – if the stars align – monetization. Here's where our software engineers, alongside their brainy buddies, the data scientists and AI maestros, really strut their stuff in the ethereal realm of the cloud. They're slicing and dicing data, extracting insights, and taking action based on those revelations. And let's not sideline our electro-mechanical engineering pals. Sure, they're crucial in making sense of the data, but in this arena, they're more like data consumers rather than the creators.


For instance, an automotive engineer might spot some quirky data that flags the need for a part redesign, or pinpoint a glitch where the software is messing up with an actuator. That's their cue to pass the baton back to the software engineer for troubleshooting. But wait, there's more – where the cloud kisses the car, we hit a classic 'edge' scenario. This is a complex dance that demands the collective intellect of multidisciplinary minds to design solutions. And here, our traditional automotive experts often hit a roadblock. For many, it's just one leap too far into uncharted territory.


Buzz Word Bingo: Cloud, Edge, and Containerization


A few years back, I found myself across the table from the CTO of a major automotive player. The stage was set for me to delve into my pet topics: cloud, edge, and containerization – all crucial tech levers, in my view, needing a holistic approach. Here’s a snippet of how that conversation unfolded:


Me, trying to gauge their stance: What's your take on the idea of a sub-cloud within the vehicle?


Them, after a moment of deer-in-the-headlights look: Oh, right... we're teaming up with AWS on that.


Me, pushing further: And your thoughts on the edge services block concept I floated by you?


Them, somewhat dismissively: Seems like a lot of hype. We're focusing on over-the-air updates.


Me, curious about their AI strategy: How about leveraging AI for your safety certification challenges?


Them, confidently: We've got NVIDIA on board for the GPUs, so we're covered there.


The dialogue continued in this vein for an hour – a predictable loop I've experienced with numerous execs. Now, I'm not recounting this to knock them down or toot my own horn. But, let's face it, these topics are bread and butter in the IT sphere across various industries. Fast forward a year, and my team clinched a multi-billion dollar deal with a Tier-1 based on these very concepts. However, they're still collecting dust on the drawing board. It's a classic case in the automotive sector – a glaring example of the big disconnect. The reason why we're not breaking any eggs here.


So, let's dissect the real conundrum with these buzzwords. Sure, the automotive sector is hopping onto the containerization bandwagon, but let's stop and ask the million-dollar question: Why? Is it to usher in a genuine era of microservices architecture, forging a seamless link between cloud and vehicle? Or is it just a fancier way to ship code? See where I'm going with this? Are we leveraging containers as mere tools, or are we eyeing them as catalysts for something bigger, something with a price tag attached?


Now, let's pivot to Edge. If you listen to the big players in cloud computing, they'll tell you that edge computing happens at the base of a cell tower. That's one slice of the pie, sure. But let's go deeper. I recall a time back at Wind River when I delivered a talk about the 'extreme' edge in automotive, diving into the nitty-gritty of weaving vehicles into a broader cloud computing tapestry. In my humble opinion, that's the edge that demands our full attention.


Out of all the global execs I've jawed with, only a couple really got it – really saw the big picture. And would you believe it, one of those visionaries ended up buying our company for $3.5B. That tells you something about where the real value lies, doesn't it?


Does SDV Really Matter?


That's the million-dollar question, isn't it? Does SDV actually matter? Let's peel back the layers on this one. For starters, let's face it – cars have been cruising along just fine for over a hundred years. If we're all about maintaining the status quo (and being safe), then sure, let's keep churning out cars the good old fashioned way and stop pouring funds into development black holes that don't really pay off for the folks holding the shares.

But wait a minute. If we're gunning to revolutionize automotive design, to seamlessly weave cars (and other modes of transport) into a safer, greener, and more scalable transport ecosystem, then we've got to shift gears on this conversation. It's not just about tweaking a few nuts and bolts. We're talking about cooking up a transportation revolution here, and that's going to need a whole smorgasbord of brainpower to crack some eggs and whip up the kind of omelet that'll make history.


Now, I get this query a lot – does SDV have anything to do with electric vehicles (EVs)? Short answer: Not directly. You can roll out EVs using the same old templates we've been applying to our gas-guzzlers, with a few tweaks for the battery tech, regenerative braking, and whatnot. The same goes for hybrids, despite the added juggling act between power sources. So, the presence or absence of SDV isn't a deal-breaker in the EV saga.


But here's the kicker: If these big-shot automakers are serious about making some serious coin off these vehicles, as they claim, SDV could be the ace up their sleeve. It's not just about getting from A to B anymore. It's about turning these rides into rolling goldmines. However, as I’ve previously stated, we don’t have the right butts in the seats to make this massive shift.


Finally, let’s talk about something that often flies under the radar with SDV, a gem at the heart of any software-defined-whatever (SDx) tech. We're talking about the power to whip up systems on-the-fly, at breakneck speed and scale, no matter what hardware is under the hood. This, my friends, is the real deal of abstraction.


Picture this: You've got your car frame, engine, and body all set (or let’s say skateboard, battery, and body for EVs). Then you slot in the compute power, load up the software, tweak it a bit, or – if we're really riding the AI wave – let the software tune itself. I can already hear my safety engineer pals slamming on the brakes, yelling, “Hold up! What about safety?” Sure, safety’s a big deal, but guess what? AI can be a game-changer in safety certification, especially when it’s artfully woven into the mix and steered by the old hands in the business.


I’ve been mulling over this a lot, and frankly, the whole safety certification shebang, across all industries, is ripe for a relook, especially through the lens of monetization. Injecting AI into this process could slash years off safety certifications, cranking up the innovation dial to warp speed. And this, right here, is where you need us mavericks from outside the box. We're the ones who can stir the pot, bringing fresh perspectives to old problems.


Closing Thoughts


So, let’s get to the heart of the matter. The emergence of Software Defined Vehicles (SDVs) is kind of like opening Pandora's box in the automotive world. On the bright side, we're stepping into a renaissance of innovation. Sophisticated software is reimagining cars as more than just hunks of metal and rubber; they're becoming smart, dynamic beasts, a stark leap from their mechanical ancestors. But here's the rub: this evolution spotlights a stark incongruity. The industry, which has long prided itself on the might of mechanical and electrical engineering, is now at a pivotal juncture. Sure, these engineers have been the driving force behind automotive marvels, but their stronghold is being challenged by the rising titan of software engineering. This field isn't just knocking at the door; it's kicking it down, becoming indispensable to automotive design and functionality. This shift isn't just a wake-up call; it's sounding the alarm for a seismic shift in how we think about building cars. As we hover on the cusp of this automotive revolution, we've got to ask ourselves: Are the traditional engineering maestros ready to tune into this new rhythm, or are we barreling towards an urgent need for a complete overhaul in our approach to car making?







6 views0 comments

Comments


bottom of page