Don’t Let The Cost of Over-The-Air Updates Skyrocket

Automakers are delivering more over-the-air (OTA) updates than ever before as they strive to keep up with customer demand and enhance their User experience. Tesla has been leading the way in this respect with one OTA update roughly every four weeks. This is only set to grow, and many automakers are beginning to catch up.

In our new 2022 automotive software survey, we found that over 40% of respondents expect each connected vehicle to receive between two and six OTA updates per year from 2025 onwards and nearly 20% expect between seven and 12 OTA updates per year. This still might be a little way behind Tesla, but it's clear that even more traditional car manufacturers are leaning on OTA updates to deliver new features, fix bugs, and even improve a vehicle's security and safety.

The challenge for automakers, however, is the cost of delivering these updates. Our recent cost consideration guide found that the cost of full image updates could be as much as $2.7 billion when you take into account the cost of data transmission, cloud storage, and dual bank memory for a single large OEM (10m+).

Even though legacy binary updates require endpoint integration costs, they can help an automaker save money. We estimate the cost of these types of updates to be around $1.8 billion per year.

However, there's another type of technology that can reduce costs even further. Line-of-Code Intelligence offers a clientless solution that creates small update files, significantly reducing costs in three key areas.

1. Eliminating the need for dual memory

Even though costs have come down in recent years, the use of dual flash memory can soon add up across thousands of vehicles. It's possible to mitigate $1.7 billion in costs using Line-of-Code Intelligence technology to create the smallest fully executable update file possible.

Where other update types overwrite previous versions and require dual memory to allow for a rollback in case of issues, a Line-of-Code update simply writes to the next free space on the memory, meaning nothing is written over, and all previous versions remain. Not only does this help save on initial hardware costs, but can extend the life of the memory by reducing write-and-erase cycles.

2. Reducing cloud and data needs

Line-of-Code Intelligence technology uses AI and advanced algorithms to make the update files six times smaller than binary diffs. This means automakers could save around $73 million in data transmission and cloud storage costs. On top of this, the smaller file size also improves transmission times, which helps get OTA updates out more quickly.

3. Clientless technology

With Line-of-Code updates, there's no need to integrate proprietary software onto every ECU. Additionally, the updates use the same file format as the original embedded ECU file (ELF, S-Rec, Intel HEX); this guarantees all ECUs can receive the necessary updates without extra integration work.

This can reduce integration costs by as much as $1.4 million but also speeds up the development and delivery process by using technology and file formats that are already integrated into the toolchain. This removes the challenge of adapting testing, production, and maintenance systems.

For vehicle manufacturers that are starting to see the benefits of software as a revenue generation tool and those who are looking to better serve their customers, OTA updates will play a large role in the future. In our new 2022 Automotive Software Survey, we found that 62% expect up to 10% additional revenue for OEMs from selling software features OTA by MY 2027.

In order to manage the costs of these updates and create a more sustainable business model, automakers should look to AI and Vehicle Software Intelligence to help solve the problem of skyrocketing costs. If you'd like to know more about how Line-of-Code Intelligence technology could help save you money, get in touch here.

Continuously Better — a Recipe for Winners

Making devices, processes, and even people continuously better is not a new idea. Case in point -- we were watching a documentary about Henry Ford last week and one of his main business principles was 'good isn't good enough', and he continuously made improvements to an already revolutionary factory floor. He implemented processes and technology -- mainly the conveyer belt - to make sure building the Model T was more seamless, bringing automotive parts to employees instead of employees going to find parts. Fast forward almost 100 years, and the same principle is executed by successful companies.

The browser wars were won because Google ensured Chrome was always being improved with seamless updates increasing speed, enhancing security and introducing new features. In the automotive world, where software is now a driving force, vehicle manufacturers are continuously making the consumer experience better with over-the-air software updates. The problem is they are not always seamless.

For example, a friend recently received a letter from his car manufacturer explaining that an update to fix the infotainment system was available. The options were to take the car to the dealer or to update the vehicle on his own. A link to instructions on how to do the update himself was in the text of the letter. Going to the noted website, he landed on a YouTube page with a video of how to do the update. When he went to his vehicle to follow the instructions -- they were totally incorrect, and the UI in the video didn't even match the UI in the vehicle. Not very seamless.

Tesla is always used as an example of how to best do seamless over-the-air updates offering new features and functions that consumers will look forward to and enjoy. On the other end of the spectrum, there are those that think updates are "Big Brotherish" and should not be allowed at all.

However, continuously better will always win. But, unlike mobile phones and laptops, 'continuously better' for the vehicle requires a great deal more effort on technology testing, quality assurance, third-party certification and regulation. The generation coming up in the world expects their vehicle to mimic their phone, and they want the same user experience. The generation building these solutions today is responsible to make sure 'continuously better' keeps the next generation safe while simultaneously meeting their expectations of personalized and satisfying user experiences.

Part 2 – Challenges of Using Software as a Revenue Generation Tool

While both new car manufacturers and traditional OEMs are embracing software updates as revenue generation tool, there could still be some bumps in the road ahead. There's no doubt that adding additional features and upgrades to a car after the initial sale can drive new revenue streams (see part one of this blog series) for carmakers, but all great opportunities come with their share of challenges.

Currently, the cost and complexity involved in overhauling legacy plated processes mean some OEMs have forms been slow to adopt OTA updates as a method of delivering feature and firmware upgrades. However, legislation, new regulations, and maintaining the user experience could present an equally sizeable barrier.

Cost

As vehicles become more sophisticated, the cost to keep them up-to-date with the latest features could easily spiral out of control. Some methods of updating a vehicle require huge amounts of data to be stored and transmitted for every update. Both full-image and binary updates could see costs run into the millions for cloud storage alone. To provide the type of updates demanded by consumers, OEMs need to look for ways to reduce these costs.

One method is through Line-of-Code Intelligence, which doesn't fully overwrite the flash storage in a vehicle. Instead, it just updates what is necessary and writes to the next available space on the chip. This can help to reduce costs (as there's less data to store and transmit), as well as improve the experience for the end-user.

User experience

As a driver, there's nothing more frustrating than jumping in your car only to find out you have to wait for the vehicle to update before you can use its core functionalities. If manufacturers are to deliver new features to a vehicle in order to increase revenue, the experience needs to be seamless.

Because, full-image and binary updates erase the previous code, the driver would need to wait while the car is being updated. In most cases, this shouldn't take too long, but it's far from convenient. Line-of-code updates are a little different, however, and allow the driver to continue on as normal with no break in how they use their vehicle. This is because the previous code isn't erased, so the old version of the software can continue to run while the update is being delivered.

Safety concerns

While many manufacturers are currently able to make updates as needed to their vehicles, some experts have safety concerns, arguing that novelty and performance features could cause problems. Even where OTA updates are delivered to improve safety, the argument is that these may not have been adequately tested in the same way they would be at the point of manufacture.

Legislation is coming into place that references OTA updates, how they're tested, and the impact of new safety features. This is on top of insurance validity concerns around changing the functionalities of a vehicle, especially when manufacturers offer free trials of different services or those on a subscription.

The UN has already established a set of rules around cybersecurity and software updates. The WP.29 rules (R155 and R156) will come into force in the EU in July 2022 and will be mandatory for all vehicles by 2024. While many of these rules surround cybersecurity, they're also focused on "providing safe and secure software updates and ensuring vehicle safety is not compromised."

As well as safety and security around software updates, WP.29 will also cover these areas:

  • Managing vehicle cyber risks
  • Securing vehicles by design to mitigate risks along the value chain
  • Detecting and responding to security incidents across a vehicle fleet

Manufacturers will need to comply with these regulations for all features delivered with the vehicle, as well as those delivered via an update. While it will take time for these regulations to come into force fully, it's important that manufacturers take steps to ensure they are fully and satisfactorily adopted.

Insurance increases

Many insurers consider new features delivered via an OTA update to be a modification to the vehicle. This could lead to an increase in insurance prices or, at worst, render the cover invalid. We all know to report modifications to our insurer, but the rules around new features delivered over the air aren't quite so clear.

Recently, UK insurer LV did a U-turn on its policies after charging Tesla owners a premium following routine software updates. It told the consumer association Which?: "We now recognize that it isn't fair to expect customers to contact us for every update, so as a result of this valid challenge, we are changing our approach."

With no existing set of rules for insurers, each will decide its own approach to these updates. This could make life difficult for consumers and could impact how car manufacturers deliver updates in the future.

Vehicle Software Intelligence as a solution

While there may be challenges ahead for OEMs, the opportunities for revenue generation are too good to ignore -- especially in this rapidly evolving market. One solution that could ease the pain of these safety and regulatory challenges is artificial intelligence, specifically Vehicle Software Intelligence. This makes the update process more straightforward for car manufacturers by minimizing the size of update files, reducing costs, and giving accurate visibility of a vehicle’s entire software system -- supporting auditing and compliance efforts.

While software-defined vehicle manufacturers are leading the way when it comes to delivering OTA updates, legacy OEMs are catching up. In fact, more than 20% of industry experts expect software sales to account for at least 10% of carmakers' sales by 2027. The road may not be as smooth as some may hope but it's the early adopters that will reap the rewards in the years to come.

Find out more about how Vehicle Software Intelligence could help your business here.

Part 1 – Software will create profitable new revenue streams for OEMs

For many years, software has been an enabler for hardware, but, increasingly, it's becoming a source of revenue for automotive manufacturers. In fact, according to McKinsey, data-driven services could create up to $1.5 trillion in additional revenue for OEMs.

Profits on a new car are ridiculously tight, with many manufacturers making just 13-21% gross profit margin (GPM) on a car sale. These margins are squeezed even tighter thanks to supply chain disruption; increased steel, energy, and logistics costs; and more R&D spend. Compare this to the software industry, where the average GPM is 72.31%, and it's no wonder OEMs are turning to software to boost margins. Many are already using over-the-air (OTA) updates to deliver new features to vehicles even after the initial sale.

One of the most well-known examples of this is Tesla offering acceleration upgrades to its vehicles with a simple update -- for a fee, of course. But there's potential here for other brands to follow suit, and many already are.

Giving more to customers

Adding new vehicle features via OTA updates will not only drive revenue to manufacturers but can improve customer satisfaction too. Drivers can add new functionality to their vehicles as need dictates or if they want to customize a used vehicle to their needs. But this is so much more than updating satellite navigation or adding new infotainment features, software gives OEMs the chance to update the functionality of the car itself through firmware updates.

This often includes small upgrades to improve the performance of a car, as in the case of Tesla. The Polestar 2 is another example, however, it gained 67 horsepower from an update to the powertrain ECU, with a retail price of $1,125. This is a fantastic indication of what we can expect in the future as OEMs begin to use software as a revenue generation tool.

Tesla is used to monetizing these upgrades and does so with great success. For example, for $10,000, you can buy the full self-driving package for your car. Tesla makes this easy for customers and simply delivers these new features through an OTA update. This, essentially, activates the existing hardware enabling drivers to make use of Tesla's full suite of autonomous capabilities.

BMW is another manufacturer that's offering more to its customers through remote updates. Owners can choose to add a range of digital services to their vehicles -- either for a one-off price or a monthly subscription. You could add active cruise control, adaptive suspension, or BMW Drive Recorder -- this automatically activates in the event of an incident but can be used to record beautiful surroundings and road-trip memories at the touch of a button if you choose.

New revenue streams

It's not just BMW, Tesla, and Polestar monetizing these updates. Stellantis recently announced a strategy that will build on existing vehicle capabilities to transform how customers interact with their vehicles -- the company is predicting this strategy to generate 20 billion euros in incremental revenue by 2030.

Stellantis CEO Carlos Tavares said: "Our electrification and software strategies will support the shift to become a sustainable mobility tech company to lead the pack, leveraging the associated business growth with over-the-air features and services and delivering the best experience to our customers."

Manufacturers have plenty of options when it comes to monetizing functionality upgrades via remote updates. One-off fees add permanent features to a vehicle but the subscription model allows OEMs to create recurring revenue streams. General Motors is already using this to add new functionality to older models. Owners of around 900,000 vehicles built from 2018 can add navigation to their infotainment system for just $15 a month.

The opportunities are endless here, especially as manufacturers look beyond their infotainment systems and to the firmware of a vehicle. Updating the features of a car -- such as safety systems, performance, or self-driving capabilities -- is where the real money lies.

OEMs aren't shy about their plans to make money from these additional features. Markus Schafer, head of research for Mercedes cars, told CAR magazine: "We're aiming for an additional 1bn euro by 2025 to be added from packages and services that we're selling over the air. Of course, we want to provide features and new experiences to our customers, but also ultimately to do additional business in the future after we've sold the vehicle. That's going to be more and more important."

For consumers, OTA updates mean personalization for their cars, allowing them to add all the features they require to an otherwise standard used car. Adding these functionalities enables OEMs to continue earning from older vehicles while promoting brand loyalty among used car buyers. While software-defined vehicle manufacturers are leading the way, legacy OEMs are catching up. In fact, more than 20% of industry experts expect software sales to account for at least 10% of carmakers' sales as early as 2027. The road may not be as smooth as some may hope but it's the early adopters that will reap the rewards in the years to come.

 

Read Part 2 of this series here

Why is hardware a major cost driver in vehicle software updates?

In the world of software updates - why are we even talking about hardware?

For in-vehicle applications, the software that runs different functions is installed directly into the vehicle (unlike cloud-based applications that we're getting accustomed to in other industries). There are as many as 100 mini computers or electric control units (ECU) in a vehicle, each consisting of both software and the hardware components that run them. When manufacturers plan the ECU components, one of the key decisions is what size memory to allocate. Chip makers price memory storage (Flash and RAM) differently based on capacity - and the price difference between the different sizes is tremendous.

To ensure the cost of a vehicle is as effective as possible, manufacturers purchase memory chips based on the size of the program file. Since flash sizes are standard and program sizes can vary, often not all memory is utilized. For example, if one program (measured in compiled binary code) is 13MB, the closest chip size is 16MB, leaving 3MB free for software image growth over time.

To conduct software updates, a sufficient amount of available space must be available on the memory drive.

Why do OTA solutions require available memory space?

There are a few different methods in the market today to send update files to vehicles. Some send the entire program all over again, and some send the delta or only the changes between the current and the new version of software. However, to assure a failsafe update process, all the different methodologies have one thing in common: you don't erase the existing version until the new one is safely installed. Verifications of the newly updated version and the enablement of a safe rollback to the previous version is required to ensure a smooth transition.

This means that at any given time, an ECU's memory must allow for both the original program as well as the complete update file simultaneously - making the need for memory greater than the original size of the program. Using our earlier example, a 13MB application taking up 81% footprint may not leave sufficient memory space for additional update files.

The initial (now legacy) OTA update methods require a full second/redundant bank of flash memory, also known as A/B memory. Whether a full image or binary diff update, it is very costly and requires creating dual partitions in the endpoint memory and doubling the available memory. In this manner, the updated software version image is written to the second partition, and if it fails, the system can revert to the previous version that is located on the first partition.

How much can a little flash drive possibly cost?

Conservative prices for 512MB NAND flash chips that might be used in domain controller architectures cost about $8 each, but a manufacturer may produce several million vehicles annually with multiple (3-4) domain controllers, TCU, Head-Units and gateways. Even in architecture with smaller discrete micro-controller ECUs that use 2, 4, 8 or 16 MB flash memory chips, they typically still cost $1 or more each, and a vehicle may have 100 ECUs that are required to be updateable. If a manufacturer builds 10 million vehicles a year that require double banks of flash memory, the additional cost could easily surpass $1.5 billion.

To read more details about potential costs of OTA updates, visit the Guidehouse Insights' Cost Consideration Guide.

Aurora Labs' additive update files - the cost-effective alternative

Aurora Labs' Line-of-Code Intelligence technology has revolutionized the way that updates files are created. Unlike existing binary diff update methods, Aurora Labs' Auto Update integrates into the development toolchain and automatically analyzes and identifies changes in the lines of code. With this intimate understanding of the software, we are able to generate the world's smallest update files (6x smaller than other binary diff methods and a fraction of the size of a full software image). Our small update files are additive update files that are written to the next free space on the existing flash memory. As many as 20 update files can be written to the existing free space avoiding the need for additional memory drives or even external memory 'boosters'.

To learn more about our OTA technology, visit the Auto Update product page.

The silver lining of complex vehicle software QA

Can you imagine manually QA'ing a system that looks like this?

We certainly cannot. And based on the recent Automotive Software Survey conducted by Strategy Analytics, neither can most industry developers:

 

The advent of technological advancements into vehicles represented by complex, interoperable and interconnected in-vehicle software systems present not only development challenges, but drastically increase the complexity of quality control.

As car makers and Tier-1 suppliers gear up to solve and support new quality control measures, we'd like to shine the light into some opportunities that rise from the growing complexities.

Opportunity: Implement continuous QA for vehicles that are on the road

As the automotive software industry follows software development trends, the pace is escalating using rapid development methodologies. The shift from the longer and phase-oriented development process to continued development is necessary to keep pace with the exponential growth of in-vehicle software to support electronic and software-led vehicles. The maturing software development methodologies from 'v development to CI/CD create opportunities for ongoing and continuous quality control that will in turn increase customer satisfaction and prolong the value of the vehicle. Survey respondents also reaffirm this notion looking at upcoming new model development.

Opportunity: Conduct proactive QA to alleviate costly recalls

As car makers implement continuous monitoring and tracking of 'on the road' vehicles, they gain the power to be proactive about threats, software friction or new bugs. This can present a huge opportunity for a major paradigm shift in the automotive industry: vehicles can detect problems before they cause system failures and remotely fix them OTA in-lieu of lengthy and costly recalls and garage base visits.

The importance of proactive actions is strongly resonating within our survey respondents as well with 93% agree with this notion:

Opportunity: Accelerate the implementation of AI-powered Vehicle Software Intelligence

As outlined in the example above, ECUs and system capabilities within a vehicle are interrelated and have direct effects on each other - necessitating a deep and current understanding of their relationships and dependencies. Utilitrends in the Automotive Software industryzing AI and dynamic monitoring, Vehicle Software Intelligence technologies will be a must-have for developing, monitoring and testing connected vehicles. These technologies will not replace existing development and QA processes and tools, but rather be an additional intelligent layer to bring static data to life and give insights at the line-of-code resolution. At Aurora Labs, we strongly believe that Vehicle Software Intelligence is the key to solving development challenges and the early onset of these technologies will help the OEMs innovate and iterate faster.

To see additional , read the full Automotive Software Survey report.

Why car manufacturers are designing the software and not just the car

It used to be the mechanical details of a vehicle that made it stand out. Buyers wanted to know who had the best engine, which four-wheel-drive system was superior, or simply which was going to be the most reliable in bad weather. While these things still matter, times change, and OEMs are looking for new ways to differentiate themselves from the competition.

In the last decade, there's been a clear shift in the automotive landscape. We're seeing new propulsion types, the rise in autonomous abilities, and a level of connectivity that feels like it's straight out of science fiction. Consumers want to know if a car will park itself, whether an over-the-air (OTA) update will make it go faster, and which new advanced driver-assistance systems will keep them safe behind the wheel.

All these innovations rely on one common factor; software -- and for new energy vehicle (NEV) startups, in-house development has been crucial from day one. Legacy manufacturers are racing to catch up.

Software as a competitive differentiator

With the average vehicle containing around 150 million lines of code, the software makes up a large part of a car's value -- dictating new features such as gesture control, self-driving abilities, and voice interaction. With the likes of Tesla and NIO leading the way with software, many other automotive OEMs are looking for ways to bring their development in-house. This would not only improve time to market but also offer clear differentiation from competitors.

These changes won't happen overnight, though, as digital transformation of this scale takes time. Our recent Automotive Software Survey showed the majority of respondents predict that 10-25% of vehicle software will be produced in-house by mass-market manufacturers in 2025.

Consumers expect OTA updates

Most buyers think of Tesla when it comes to OTA updates for a good reason. The electric-only manufacturer has been building these capabilities into their cars since the launch of the Model S in 2012. Other automakers have struggled to keep up, though most now offer some form of basic OTA updates.

What is still setting NEV companies, such as Tesla and NIO, apart is the type of updates they offer. Most manufacturers can update the software on the infotainment systems but those leading the charge can also administer OTA updates to the safety-critical systems. This means being able to make adjustments and upgrades to more complex systems such as braking, steering and ADAS. Legacy manufacturers will struggle to do more than update the navigation and infotainment systems with their current development processes and OTA update solutions.

Data from Statista shows the value of the worldwide OTA update market could be as much as $7.5 billion by 2025, meaning it's not an area automakers can ignore. To keep up with consumer demand and not be left standing by NEV powerhouses, OEMs are looking for ways to quickly increase the capabilities of their OTA updates and launch new features. Bringing everything in-house is the clear solution but this will take time, meaning OEMs will continue to work with tier-one suppliers.

Managing the transformation

Using these suppliers is still necessary for most manufacturers but the benefits of in-house software development can't be ignored. It can help keep costs down, fast-track delivery, and protect against cyber vulnerabilities but there's a solution for OEMs who still need to outsource some elements of their software development: Vehicle Software Intelligence.

The key to working with suppliers is visibility. It's important to understand the bigger picture of inter-dependence and operability between elements developed by different vendors. Aurora Labs' Vehicle Software Intelligence is an AI layer that is used by OEMs in their software development efforts and the way they manage suppliers.

One challenge, for example, is software dependencies. When OEMs rely on third parties for their development needs, it's easy to lose sight of how different systems hang together. With a Line-of-Code Intelligence solution, manufacturers to get a better view of the system as a whole. This allows developers to keep an eye on the thousands of inter-related functions and capabilities to better understand the potential effects of new features and conduct OTA updates with confidence.

With 77% of respondents to our 2021 Automotive Software Survey stating that the trend towards in-house software will increase, it's clear that automakers have some work to do. The development of software needs to be treated as a strategic move by OEMs that want to stand out as the demand for software-defined vehicles continues to grow.

Want to learn how to apply Vehicle Software Intelligence to your software? Contact us.

At the Epicenter of the Unknown

For 10 years, I have been writing about and taking the bullhorn to the mountain to talk about automotive software and the benefits of over-the-air updates. For three years, I have been writing about and taking the bullhorn to the mountain to talk about automotive software and the benefits of validating what happens when there is an over-the-air update.

This week, I experienced what happens when an update is not validated and found myself at the epicenter of the unknown. I have a swanky new 2021 SUV. This model is no longer a boxy vehicle like previous models - it is sleek and fun and has many of the infotainment, ADAS and connectivity features we talk about on a daily basis in the automotive industry.

When I first bought the car, I could say, "Call Mike," and Mike was soon on the line. Now, I say "Call Mike," and I get the response - "Ok, let me help you with that. I need some more information. Look at the notification on your device."

"Looking at my device," forces distracted driving and is obviously not recommended. This prompt goes against every goal of bringing voice assistance into the car. I went to an online consumer OEM support group and read posts noting that this problem started in November 2021. With yes - an over-the-air update.

I'm sure the update did fix some things - or add some things - I don't know. I do know that the update screwed up my ability to call out by contact name (I can call out by dictating the phone number, but out of my 210 contacts, I know three phone numbers by heart.)

So, after going through many menus, I finally went to the dealer for help.

I met with a super nice support person. He tried - but his conclusion was that it was an Android Auto and phone problem and I had to go to AT&T.

I went to the AT&T store and met with a super nice sales person. He told me he wasn't certified to help me - he cannot give advice or guidance for anything in the car for liability reasons. He did give me a phone number for the AT&T Advanced Technology Group.

I called the AT&T Advanced Technology Group and another really nice support person told me that her group only works on networking issues to the car - hotspots and things. This AT&T person told me I had to talk to the car manufacturer and sent me to a really nice support person at the OEM who also told me I was again not talking to the right group and he forwarded me to another support group within the OEM.

Here is the kicker - I don't know if the next support person is really nice. My next conversation was with a phone recording repeating, "My name is Joe. I can't hear you. Please call back later."

I do love my new swanky, new SUV. This is my third purchase from this OEM.

I also come from phone company parents - so I'm sure the phone company helped to put me through college.

All of this really nice support and sales people are doing the best they can with the information they have.

I know that we are in the early days of 100 million lines of automotive software code. I also know that validating software behavior throughout the entire car resulting from an over-the-air update is paramount and that the really nice sales and support people from both the automotive companies and the service providers - need to be educated on how to help consumers navigate to success.

For now, I am still at the Epicenter of the Unknown. Please comment if you have any insights or fixes to this 'call by contact name' problem.

Could software be the answer to the chip shortage?

A global chip shortage, prompted by the COVID-19 pandemic, is continuing to affect automakers around the world. As a result, many OEMs have had to delay new models and slow down production. However, there are positives to this challenge as it could force the industry to innovate in new ways.

Software could be one such solution and this isn't the first time manufacturers have looked to developers to solve a hardware problem. In 2014, Tesla announced a recall of 29,000 of its charging adapters over fire concerns. This wasn't a recall in the traditional sense, though, and Tesla was able to fix a problem with the charger's electrical resistance heating by rolling out an over-the-air (OTA) update to its vehicles.

Can an update solve a chip shortage?

Software-defined vehicle manufacturers are well placed to look for ways to solve the current chip shortage and, once again, Tesla is leading the charge. It's already exploring alternative chips and is rewriting its vehicle software to support that.

"We were able to substitute alternative chips, and then write the firmware in a matter of weeks," Musk said during an earnings call in July last year. "It's not just a matter of swapping out a chip; you also have to rewrite the software."

A new way of delivering OTA updates

Tesla's software mindset allows it to weather the challenges facing the automotive industry but all OEMs have the chance to learn from this. The level of technology in a car requires millions of lines of code and each OTA update needs additional storage on top of this. Reducing the memory needed for each update could reduce the pain of the current chip crisis for automakers.

This can be achieved with Line-of-Code OTA updates.

Legacy update solutions such as full-image updates or binary diff updates require creating dual partitions in the endpoint memory while also doubling the storage available. Using these methods, manufacturers update the software to the second partition so that, if it fails, it can revert to the previous version that's still stored on the chipset.

This not only requires double the amount of NAND chip memory to accommodate these updates but if an update fails, it could have a knock-on effect across other ECUs in a vehicle that may need to roll back two or three versions - something that's not possible using full-image or binary updates.

Line-of-Code updates, however, don't require this extra memory to ensure that the ECU is both updatable and fully fail-safe. Using this technology could ease some of the problems OEMs are facing from the current chip shortage - by requiring fewer flash memory banks while still balancing the need for safety and user experience.

This method writes the fully executable update file to the next free space on the flash memory without deleting previous versions. Not only does this method take up much less space than legacy updates but it means there's zero downtime for the user and, should an update fail, it's easy to revert back to a previous version.

It's clear that software has the chance to solve hardware problems but challenges such as the current chip shortage present new opportunities for innovation. Vehicle Software Intelligence and Line-of-Code updates may help ease the pain of a short-term issue but can also support OEMs in revenue generation, help them create an improved user experience, reduce costs, and offer full visibility into the co-dependencies present in a modern vehicle.

If you're interested in finding out more about Line-of-Code updates, take a look at our cost consideration guide for OTA updates.

Will the automotive industry adopt the Apple vertical integration approach or the Samsung collaborative ecosystem approach?

Before software started to dominate vehicle production, the balance in the automotive ecosystem was clear and well known: Tier-1s supply parts, mechanics, hardware and software, whereas OEMs own the design, assembly and marketing. Today, and looking forward, as software is starting to supersede mechanics in vehicles, and the core of the vehicle is changing from internal combustion to electric, the balance in the automotive ecosystem is also changing.

The shakeup in roles does not necessarily mean some parties become obsolete, but rather that all parties have the opportunity to collaborate for mutual growth.

Let's take a look at trends in key players of the automotive ecosystem:

Chipset providers are showing signs of growing their offering from hardware only to include software and services. For example, Mobileye's acquisition of Moovit is a clear move to a full robotaxi service. Qualcomm's recent acquisition of Veoneer signals the next phase of their complete software and chip platform for driver-assistance systems.

Tier 1 suppliers are displaying leadership in software development and innovation to continuously bring value and skills to OEMs. Examples include Continental's purchase of Argus Cyber Security in order to strengthen and enhance its capabilities in automotive cyber security and LG's similar venture with the acquisition of Cybellum.

Aurora Labs recently collaborated with Strategy Insights to conduct an Automotive Software survey where we polled over 150 industry leaders for insights on trends in automotive software. Based on the results from this survey, OEMs are estimated to move between 10-25% of software development in-house by 2025. (See complete survey report for more details)

This shift is very logical - as software becomes a differentiator and an integral piece of the vehicle, car makers must have competencies and transparency into the code they deploy in their vehicles.

Car manufacturers can opt to move in two directions: the 'Apple' or 'Tesla' method of complete end-to-end vertical integration and control, or for a collaborative ecosystem. It is our belief that the industry will move towards the latter.

If we look at the mobile phone industry example, although Apple has shown tremendous success with its vertical integration strategy, other companies such as Samsung have also experienced great profits through a thriving ecosystem by utilizing Android operating systems and multiple hardware suppliers.

Based on the 2021 Automotive Software Survey, there is no decisive answer as to which player today has the most influence over software development, a trend that signals a period of upcoming changes.

Cooperation and collaboration within an ecosystem drive greater innovation, lowering of costs and increase in revenue. The automotive industry currently stands at the brink of such growth, which will be driven by all participants. The new mode of operation has room for each party to evolve and work in cooperation, rather than compete with each other.

Naturally, there may be a period of unrest during the transition to the new era, with each player in the ecosystem attempting to own a bigger piece of the value chain, but this is expected to settle as the new lines are drawn.

This evolution will require more than collaboration, but will also be driven by regulation and technological advancements. Standardization and validation throughout complex and interrelated systems are being worked on around the globe. And close to our hearts at Aurora Labs, AI-powered software intelligence technologies will be the enablers for transparency, evidence, authentication, testing and updating of systems supplied by different players in the ecosystem.

To learn more about trends in the automotive software industry, read the 2021 Automotive Software Survey Report.