Comma.ai: Solving Self-Driving Cars — A Startup Investment Thesis by Mateen Asghari and John Gaidimas

Introduction to comma.ai

Comma.ai is a self-driving car startup that sells hardware and AI software to allow pre-existing cars to become self-driving. Based out of San Diego, CA, comma.ai currently sells a kit for $1000 that allows many of the most popular cars to gain self-driving features. Comma.ai’s advanced machine learning models use image processing and the many pre-existing sensors and electronics in most modern vehicles to determine a car’s surroundings and safely steer the car. Ultimately, comma.ai aims to allow their current system to drive completely autonomously without the need for human oversight.

In April of 2018, comma.ai received an additional $5 million in a Series A funding round, bringing their total funding to $8.1 million. Although comma.ai did not announce who led their Series A, comma.ai’s seed round of $3.1 million was led by Andreesen Horowitz. Comma.ai’s founder and former CEO, George Hotz, has claimed that comma.ai hopes to avoid raising any additional funding. Comma.ai’s Series A funding round gave them a valuation of approximately $50 million.

Meet the Key Members of comma.ai

George Hotz

Founder and former CEO, currently serves as Head of Research Team. Hotz started comma.ai in September of 2015 in order to make autonomous vehicle technology more accessible and affordable using existing vehicles that have the capacity for autonomy. Hotz’s claim to fame began when he was 17 years old and was the first to hack the iPhone to allow it to be used with other wireless carriers, instead of exclusively AT&T. At 21, he hacked and reverse engineered the PlayStation 3 and was subsequently sued by Sony, further thrusting his name into the public domain. At 25 he was approached by Tesla and offered $12 million by Elon Musk to replace a component of the driverless technology used in Tesla vehicles with a better system. Holtz declined this offer, opting to narrow his focus to building comma.ai. In September of 2018, three years after comma.ai’s launch, Hotz stepped down from his position as CEO and appointed Riccardo Biasini to take his place. Hotz also left the company briefly in March of 2019, only to return in May of the same year. Hotz has also applied his coding acumen while working for Facebook and Google, specializing in AI algorithms.

Riccardo Biasini

Current CEO of comma.ai, formerly served as VP of Quality. Biasini originates from Italy where he received his Master’s Degree in Automotive Engineering from the University of Pisa in 2010. Following graduate school, Biasini moved to the U.S. From 2011–2016 Biasini was employed at Tesla where he was in charge of the Traffic Aware Cruise Control in Tesla cars along with other safety and control features. In 2016, he joined comma.ai where he developed the automated spatial controls for the company’s first driverless car. He was then promoted to VP of Quality in 2017 and in the following year he rose to CEO. Biasini argues that today’s car manufacturers and self driving companies are missing the most effective solution of widespread driverless technology: to implement self driving technology on cars consumers already own. His mission is to provide exactly this technology that will make driverless technology readily available to most drivers.

At the moment, comma.ai is composed of 13 highly-trained and extraordinarily-skilled employees in order to minimize complexity and deliver the best results.

We firmly believe this team is dedicated to seeing its company succeed in its mission of making driverless technology widely available. George Holtz, the founder of comma.ai and original CEO has proven his professional skills in the realm of software development and AI, and he still plays a prominent, positive role in the company. He disclosed that his personal mission is to “solve AI,” a goal he wishes to fulfill with comma.ai. The potential success of the company is aided by a handful of employees who are immensely proficient in the skills necessary to make comma.ai a dominant player in the domain of self-driving companies. Additionally, Riccardo Biasini, the current CEO, seems completely committed to the company as apparent in his frequent blog posts detailing the current state of the startup and where it is headed. Admittedly, some may have reservations with the company’s founder, George Holtz, due his sudden abdication from his position of CEO and brief leave of absence from the startup. At its face, this does signal questions of loyalty to the startup. However, Holtz clarified that stepping down was his sole decision because he believed he was not the best CEO for comma.ai due to his problems with execution. He put Biasini in charge convinced he was the best alternative for the company, shedding personal hubris in favor of comma.ai’s best interests. As a whole, the team at comma.ai is passionate about seeing the ubiquitous appliance of driverless technology and works to be at the forefront of this phenomena.

Comma.ai’s market size and strategy

If comma.ai’s self-driving ambitions are realized, they will be poised to drastically change the global transportation and logistics market. This is one of the largest worldwide industries. According to estimates from Plunkett Research, the transportation and logistics industry had $5.1 trillion in global revenues during 2018. The United States’ share of this revenue was estimated to be in excess of $1.2 trillion. This entire industry would be drastically changed with comma.ai’s self-driving technology. It is obvious that comma.ai has a large market to expand into with their technology.

At the moment, comma.ai’s focus is on consumer facing vehicles, but it is reasonable to assume that their self-driving technology could be easily adapted to expand into commercial trucking and delivery markets. The market for consumer and commercial self-driving vehicles is a huge area of growth. Research from Frost & Sullivan estimates that the autonomous driving market is expected to grow to $173.15 billion by 2030, and comma.ai is in an ideal position to capitalize on this fledgling market.

Comma.ai has separated themselves from many of their competitors by focusing on outfitting existing cars with self-driving capabilities instead of attempting to build a self-driving car from the ground up. This focus on affordability and support of legacy systems will provide comma.ai with a significant advantage in the commercial transportation market where every dollar counts. Many companies will be reluctant to replace their entire existing fleet with expensive autonomous vehicles, so comma.ai’s product will be quite appealing to customers in the commercial transportation market.

Although many of comma.ai’s competitors have much more funding, comma.ai has shied away from large funding rounds or partnerships with larger companies. Although many advertising materials would like to suggest that autonomous vehicles will enter the market in the next few years, most research firms estimate that these fully autonomous vehicles are at least 10 years away from entering the consumer market. As comma.ai’s CEO, Riccardo Biasini, explains “The problem, this is going to take at least 10 years, and we’ll have to go through a harsh winter. And when it comes, I hope we can survive it”. Hotz argues that many of comma.ai’s competitors are burning too much of their funding to survive this decade long period before self-driving cars have matured enough for these companies to start selling their cars to consumers. Comma.ai’s approach has been to gain revenue through selling their self-driving kits to consumers which reduces their burn rate. Ultimately, this strategy will allow comma.ai to outlast many of their competitors who are losing millions every month.

As of 2017, comma.ai’s plan to expand into the self-driving market has been to continually improve their machine learning models until they are able to ship a level 3 self-driving system. In a level 3 system, human drivers will not be necessary in certain stretches of the road, meaning that someone would be able to sleep through part of their commute. Once comma.ai has accomplished this, their plan is to have an insurance provider underwrite a car insurance policy for drivers with comma.ai’s technology. Because comma.ai’s system will theoretically be safer than a human driver, insurance companies will incentivize consumers to buy comma.ai’s technology to receive a less expensive insurance policy.

Although comma.ai’s strategy will allow them to easily expand into the consumer market, this approach is a waste of the commercial potential for their technology. Selling a self-driving kit to consumers has much less potential for growth and profit than partnering with global shipping companies to revolutionize the commercial shipping industry. Comma.ai’s unwillingness to partner with larger car companies will reduce their potential to grow into the mass consumer market, as many consumers will be uncomfortable installing this system into their own vehicles.

Current competition

The driverless technology market is a very modern, technologically forward sector. Due to this and the technology’s potential to uproot the traditional transportation paradigm, there are many startups competing in the market as well as traditional car manufacturers shifting their focus in this area.

Headquarters: Foster City, CA; Funding: $955 million; Founding date: July, 2014

Zoox is a startup focused on creating self-driving technology for cars. Specifically, it aims to make a sort of autonomous taxi system that make mobility easier and more social. Unlike comma.ai, Zoox designs cars from the ground up retrofitted specially for the purpose of autonomy. So far, the technology they have created thrives in complex environments like cities and is effective in difficult situations such as navigating right turns on reds and understanding when there are double parked cars. Zoox notably hired Aicha Evans, former Chief Strategy Officer at Intel, earlier this year which adds impressive weight to their competition against comma.ai. Moreover, a few Tesla employees defected to Zoox recently which led to a suit against the company in March of this year led by Tesla. With Zoox’s enormous amount of funding, solid leadership, and competent technology they pose as a significant competitor against others in its industry.

Headquarters: Mountain View, CA; Founding date: December, 2016

Waymo, formerly called the Google self-driving car project, is a self-driving technology company with the goal of making driverless vehicles safe and simple to use. They are a subsidiary of Alphabet Inc. and they’re valued at about $105 billion, not surprising as Alphabet essentially has a limitless amount of resources and capital for Waymo to take advantage of. Waymo is one of the pioneers in the field of autonomous driving seeing as they achieved the world’s first fully self-driving trip on open, public roads in a specialized car without pedals or a steering wheel. In April of 2017, Waymo started a pilot self-driving service in Phoenix, Arizona with 1,000 of their driverless cars. Also on October 30, 2018 Waymo was the first self-driving technology company to receive a permit from the California Department of Motor Vehicles which allows the company to operate fully driverless cars on public roads in the day or night.

Headquarters: Palo Alto, CA; Founding date: July, 2003; Revenue: 21.461 billion

Tesla needs no formal introduction as the company has dominated the electric vehicle and now the autonomous driving markets. Tesla began early in the development of self-driving technology as in 2014 all Tesla vehicles were equipped with semi-autonomous driver assist. Today, many features of autonomy for driver assistance is prevalent in their cars: adaptive cruise control, lane departure warning, Autopark, Autosteer, and summoning the car from its parking place. Tesla seems to be further ahead in achieving full autonomy compared to its competitors but the date at which this will occur is not so clear. Tesla’s revenue last year was $21.461 billion though it is losing money with the company’s net income just shy of -$1 billion USD.

Comma.ai’s product

At the moment, comma.ai’s main product is the EON DevKit that is sold as a dashcam. The EON DevKit includes a OnePlus 3T android smartphone with comma.ai’s dashcam software preinstalled as well as several necessary cables and mounting hardware. For an additional $400, customers can also purchase a comma.ai Car Harness which includes a high precision GPS and an OBD II reader which allows comma.ai’s software to gain access to a car’s internal computer. Both the EON DevKit and Car Harness are required to use all of comma.ai’s self-driving features.

The EON DevKit and Car Harness tap into a car’s existing computers and sensors. This data is then communicated to a smartphone that is mounted on the car’s windshield, and this phone feeds its camera data and the car’s sensor data into a machine learning algorithm that determines the car’s surroundings and chooses the best path for the car. The phone then sends this data back to the car’s computers to dictate how the wheel should be turned and apply the brake as necessary.

Although comma.ai’s current products have been primarily used by technology enthusiasts, they have continued to make their product more user friendly to appeal to the everyday consumer. Additionally, reviews from early adopters of the EON DevKit have been largely positive. As with any cutting edge technology, comma.ai’s software and hardware is far from perfect, but they have continually made progress to both their self-driving software as well as the hardware that is used to deploy it. While comma.ai’s technology is quite promising, it remains unseen if comma.ai, or any company, is truly capable of developing a completely autonomous vehicle.

The largest current issue with comma.ai’s product is that they are restricted from selling their EON DevKit with their self-driving software included. Before the launch of EON DevKit’s predecessor, the comma ONE, George Hotz received a letter from the National Highway Traffic Safety Administration (NHTSA) warning him against releasing an aftermarket self-driving upgrade. Following this letter, Hotz scrapped the release of the comma ONE, and instead released the EON DevKit which is advertised as a dashcam to avoid the regulatory issues faced by the comma ONE. The EON DevKit does not ship with any self-driving software, and consumers are responsible for downloading comma.ai’s self-driving software from GitHub. In this way, comma.ai avoids the regulatory issues and liability that comes with selling a self-driving car upgrade.

Although this method allows comma.ai to avoid regulation and liability, the loophole they exploit alienates consumers who are not technically inclined or unwilling to install such an experimental and unregulated system into their car. This will present a major issue when comma.ai moves to expand their product into the everyday consumer’s car.

Metrics

  • 13 full time employees
  • 30,000,000+ miles driven
  • Over 5,000 fleet members
  • About 400 Openpilot updates
  • For every user that purchases their product, comma.ai makes $400 profit
Last two years of revenue at comma.ai; current monthly revenue at ~$150,000

Conclusion and Final Remarks

It is clear there is an extraordinary level of demand for self-driving technology in the near future. Comma.ai is a practical competitor in this field as they are one of the few in this industry who have a product out for consumers to purchase and use — the other notable company being Tesla. Estimates for self-driving technology being standard in all vehicles have varied throughout the years, though now it looks to be about 5 to 10 years on the horizon. Compared to its competition, comma.ai is relatively underfunded and seems to have gained less traction in the startup world. However, the intelligence of the team and universality of the product combine to create real potential.

Compared to their competitors, comma.ai has several advantages. Primarily, comma.ai has been able to outsource a significant portion of the cost of developing their technology to consumers who pay to use their EON DevKit. Unlike most of comma.ai’s competitors who must pay for the test fleet, test drivers, and all the equipment maintenance, comma.ai relies on early adopters to gather testing miles to train their machine learning algorithm. This strategy has allowed comma.ai to greatly reduce the costs of developing their self-driving technology. Additionally, comma.ai has actually shipped a real self-driving product that is actively being used by consumers. With the exception of Tesla, there is no other autonomous vehicle company who has opened up their technology to the mass consumer market. The experience comma.ai has gained with shipping a real product to real consumers should give them an advantage as their technology continues to grow.

However, comma.ai faces many difficulties and disadvantages when compared to their numerous competitors. The self-driving startup business is saturated by many of well-funded startups that have raised billions of dollars. Comma.ai will struggle to compete with these much larger and better funded startups. While comma.ai has an innovative strategy that will vastly reduce their development costs, they completely lack the advantages of a larger, better funded startup. Comma.ai cannot afford to spend money on advertising or a legal department, and as a result comma.ai remains a relatively unknown brand that sells their product through a legal loophole. This lack of regulatory approval and oversight will make it challenging for comma.ai to expand in the consumer market, and their lack of funding will make it difficult to deal with future regulatory issues. Because they are forced to sell their product as a dashcam instead of a self-driving system, comma.ai’s EON DevKit is not consumer friendly. It will be impossible for the everyday consumer to install this into a car, and the technical skills required to actually use comma.ai’s product vastly reduces their potential market. While comma.ai could certainly raise more funding to help tackle these issues, as things stand right now comma.ai is ill-equipped to effectively compete in the cutting edge self-driving startup industry.

The question of whether we would invest in this company still stands. To be frank, while we think their technology and team’s dynamic are impressive we would not invest in comma.ai after doing extensive research of the startup and looking at the numbers. The market for self-driving vehicles is crowded with big-name and massively-funded companies and comma.ai is just not a potent player in the business at the moment. This stems from their lack of funding, $8.1 million compared to Zoox’s $995 million, and limiting small team of 13 employees. This does not allow them for a lot of upward mobility and expansion and is a glaring concern for investors who look for big profits soon. Moreover, the future of comma.ai’s product does not seem to be promising if all vehicles in the future are built from their conception to be self-driving. At that point, only their technology would be valuable and even then large companies like Tesla and Waymo may have them beat. Despite our unwillingness to invest into the company, we do believe what they are doing is impressive and that they have potential. No other company, besides Tesla, has a product on the ground running for consumers to purchase that allows self-driving to be present on their commute. Moreover, the team at comma.ai is immensely talented and capable of great technological feats in self-driving applications. Though we would like to see comma.ai become exceptionally successful and believe they have a lot of room to grow, their current disadvantages loom too large for a serious investment to be feasible.

Disclaimer: All information in this report has been aggregated through numerous, public news reports/articles or comma.ai’s website itself. Any misrepresentation of the company is purely accidental. BV is more than willing to correct any mistakes.

--

--

--

We are UCLA’s premier venture capital organization, cultivating leaders in venture capital, technology, and entrepreneurship.

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

4 challenges of cross-border shipping & how to avoid them

What would Google do?

Google home page as at 2002

How Rep Links Energize Market Action

Achieve Prosperity And Success When You Do These 7 Things (Nobody Ever Told You About)

How to use InDriver in Colombia (alternative to Uber)

3 Types of Partnerhip Business

We live in the GroupOn Economy

Brand & Branch Issues Public Comment on Proposed California Cannabis Regulation

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Bruin Ventures

Bruin Ventures

We are UCLA’s premier venture capital organization, cultivating leaders in venture capital, technology, and entrepreneurship.

More from Medium

Man-Made Systems vs. Mind-Made Systems

2022 Predictions

Analog compute for Edge AI: the Holy Grail or a road to nowhere?

Ethics of Connected Automated Vehicles