Machine Intelligence for the Innovators of Tomorrow - BMW i Ventures’ investment in Graphcore

1_Graphcore Wordmark.png
Source: Graphcore

Source: Graphcore

Today, we’re thrilled to announce our investment in Graphcore and its two co-founders, Nigel Toon and Simon Knowles. Graphcore secures the lead in the global AI chip race with $200 million in new capital from leading financial and strategic investors including Atomico, BMW i Ventures, Microsoft, Merian Chrysalis, Pitango, Sequoia and Sofina.

Graphcore is a UK based private company founded in 2016 developing an intelligence processor unit (IPU) that can improve the performance of machine intelligence training and inference by 10x to 100x compared to currently available solutions. The company is in a stage of rapid global growth, tripling the size of the team and opening new offices in London, Palo Alto and Beijing in 2018.


AI is manifesting itself at a phenomenal rate driven by deep/machine learning applications such as speech and image recognition, video surveillance, customer service, network monitoring and automotive. The annual worldwide AI software revenue will grow from $5.4bn in 2017 to $108.8bn by 2025 according to the Tractica Artificial Intelligence Market Forecast. The overall AI compute today is estimated to account for 1-2% of public cloud service sales growing to be about 30% - 40% of the revenue by 2025 [Tractica]. 

Driven by AI software compute needs, the demand for specialized hardware accelerators is imminent. The global AI chip data center revenue currently generated by GPUs, CPUs and FPGAs (Field Programmable Gate Arrays) amounts to over $4bn. It is expected that this volume will expand by 2025 to reach more than $11bn (BMW i Ventures estimation), driven mostly by sales of purpose-built AI chips, such as Graphcore’s IPU, posing a threat to Nvidia’s dominance.

The market is expected to double until 2021, so now is the time to release products in order to gain market share. 2019 and 2020 are likely the years when a ramp-up in deep learning chipset volumes will take place and “winners” will begin to emerge.

A “winner” in the chipset market can rise very, very rapidly. Looking back to the year 2000, when Intel launched a data center offensive with its x86 CPU architecture, there was a clear dominant technology in the market for server CPUs - IBM’s family of processors called “Power” - owning 40%+ market share, corresponding to more than $2bn of annual sales. The other leading technology was provided by Sun Microsystems, called “Sparc”.

Jumping to 2018, “Power” and “Sparc” are together holding less than 5% of the data center CPU market while Intel’s x86 processor family has emerged as the clear leader (s. Intel x86 data center market share development below).

For reference, there is also a distinct dominant technology in 2018 - Nvidia’s family of GPUs - owning 70%+ of the data center AI chip market, corresponding to approximately $3bn of sales expected in 2018. From 2000 to 2005, Intel grew at a phenomenal rate from about 0% to 57% market share for data centers with a revenue totaling well over $1bn [BMW i Ventures estimation].  While Graphcore’s story may unfold with even higher velocity, due to the unique characteristics of the much faster growing AI chip market today, Intel’s path shows how quickly the adoption of a new architecture can evolve in the processor world.


A decade ago, the overall CPU data center market was growing at a much slower pace than today’s AI chip market, which meant that high sales growth also translated to winning high market share. While we may not observe a gain in market share at such a rate, we expect a similar pattern to develop in the market for purpose-built AI processors, with Graphcore well positioned to lead with the exceptional combination of superior technology, team and timing.


We believe Graphcore has the right team and product at the right time to succeed in an emerging market. The company has developed a novel AI accelerator chip (IPU) that is tailored to enable the machine intelligence of tomorrow. Unlike currently available CPUs, GPUs and FPGAs, Graphcore’s IPU is purpose-built for machine/deep learning algorithms and delivers an order of magnitude better performance. IPUs provide more compute per watt than CPUs and GPUs, while they have also been designed to be extremely efficient at mini-batch training.

Source: Graphcore

Source: Graphcore

An important difference of Graphcore’s IPU is its innovative approach to memory. Computing architectures have always relied on RAM (Random-Access Memory) for program storage. This means that the variables for computation have to be fetched from the external RAM, which is both very energy intensive and is also making the memory interface between the processor and the RAM the bottleneck limiting the overall system performance.  Graphcore’s team decided to store the variables (i.e. the weights of a neural net) on the processor itself and not in the external RAM. This results in 100x+ memory bandwidth increase compared to traditional architectures.


Source: Graphcore

Source: Graphcore

CPUs and GPUs typically have 10’s to 100’s of separate processor cores. Graphcore’s first IPU processor, Colossus, has 1,216 independent processor cores on each chip. The low power consumption of the IPU allows fitting 2 IPU chips onto a single 300W PCIe card resulting in over 2,400 independent processor cores, capable to operate with more than 14,000 independent IPU program threads that are all working in parallel with minimal latency because the knowledge models are held inside in-processor memory.

To make use of these levels of data-parallelism, Graphcore has developed its proprietary software called Poplar™. It determines how the thousands of individual processors on the chip communicate with each other, making sure that data is moved across the chip most efficiently and at the right time, utilizing all available processors. This leads to significant performance improvements, enabling Graphcore to achieve processing speed gains of 10x or more when compared to today’s highest performance GPUs.

Our Graphcore Investment

The market is driven by the increasing number of artificial intelligence and machine learning applications while being limited by the performance boundaries of traditional compute architectures. This causes a large demand for scalable & optimized processing units. Being first to market with a production-ready AI chip and a unique, future-proof design, Graphcore offers an unparalleled opportunity to invest in the potential winner in that space.

So far there are limited high-performance options for central AI computation, especially when looking for companies that have systems ready for evaluation and deployment on the market. Graphcore is selling IPUs to data centers & cloud providers, while the company also has a strong product pipeline aiming towards smaller structures and automotive solutions. The combination of using Graphcore’s IPU in a data center and the possibility to utilize the same architecture in a vehicle is very attractive. In the data center, the IPUs can be used for simulation and training while the same architecture (certified for automotive) can be used in the car for inference, resulting in quicker turnaround times and less complexity during the development phase. Furthermore, the ability to run different kinds of neural nets with equally high performance on an IPU is beneficial: a flexible hardware allows innovating on the algorithms while taking a hardware design decision earlier.

This latest round of funding will allow Graphcore to execute on its product roadmap, accelerate scaling and expand the company’s international footprint. It is a further step towards fulfilling the team’s ambition to build an independent global technology company, focused on the new and fast-growing machine intelligence market.

Final Remarks

The race for the dominant architecture in the AI chip market has just begun and there is no clear winner yet. Graphcore is first to market with a production-ready dedicated chipset for today’s and tomorrow’s AI computing needs.

Graphcore’s chipsets are computationally powerful and optimized for machine learning, allowing for high throughput at a very low latency. With its versatility and flexibility Graphcore’s IPU – which supports multiple machine learning techniques with high efficiency – is well-suited for a wide variety of applications from intelligent assistants to self-driving vehicles. It’s equally good at training and inference, allowing the use in a data center as well as in a vehicle.

Graphcore has the potential to become the winner of that space. We look forward to supporting Nigel, Simon and the Graphcore team in building a major global technology company that can help innovators in AI create the next generation of machine intelligence.

Graphcore’s co-founders Nigel Toon & Simon Knowles

Graphcore’s co-founders Nigel Toon & Simon Knowles

P.S: Special thanks to our interns Marco Linner and Marie Tai for the help with the diligence.



Protecting the world’s most critical networks: BMW i Ventures investment in Claroty

Michael Dell said "Ideas are commodity. Execution of them is not."  A great example of alignment of both ideas and execution is Claroty with its innovative industrial cybersecurity platform, go-to market partnership strategy and unique team of industrial automation and cybersecurity experts. BWW i Ventures invests in cutting-edge companies that are important to the automotive and industrial sectors, and we are excited to announce our investment in Claroty ( – a company securing the world’s most critical industrial networks.

Operational Technology, or OT, describes hardware and software (i.e. programmable logic controllers (PLC) and industrial automation systems) that detects or causes a change through the direct monitoring and/or control of physical devices, processes and events in the manufacturing and other operational processes. This infrastructure is also known as Industrial Control Systems (ICS) and the terms OT and ICS are used synonymously.

Fifty years ago, computer security meant to physically protect computing devices. Ever since Bob Thomas programmed the first computer worm as a proof-of-concept in 1971, threats do not merely exist in the physical realm, but in the digital world as well. With the IT revolution, both the means of attack as well as the need for new defenses have changed dramatically. In the OT realm, the recognition of cyber-risk developed more slowly; cybersecurity protection was often viewed as unnecessary because most OT networks were not connected to the business network, the internet or other outside networks. But this too changed dramatically as industrial operators embraced automation and connectivity to increase efficiency and production.

Technological trends like smart factories, robotics, and automation or Industrial Internet of Things (IIOT) rely on connectivity to and from the manufacturing site. Data acquisition and control allow for new efficiency milestones and insights, but also increase the number of attack vectors into industrial systems. Therefore, while the industrial internet of things (IIOT) offers many advantages, it also opens OT networks to external connections and exposing them to new types of risk.

The growing threat of advanced cyber-attacks on critical infrastructure and industrial control systems presents a unique challenge for organizations. According to a survey by Russian security giant, Kaspersky Lab, 28% of industrial companies surveyed have faced targeted attacks in the last 12 months. Additionally, around 48% of companies in the industrial sector stated that there's insufficient insight into the threats specifically faced by their business. As a result of the heightened awareness of cybersecurity threats and potential, revenue and customer impacting damages, OT-security budgets are growing.


The estimated current market size of Industrial Control System (ICS) security market is still small with $450M (BMW i Ventures estimation), but rapidly growing with 55% CAGR until 2022, reaching $2.5B in 2022.

Due to the sensitive nature of operational technology, sales cycles are typically 6-18 months with small initial average order values between $200-500K. These initial sales often grow into multimillion-dollar accounts as large customers make additional purchases. Once the initial contract is won, and positive results are confirmed, there are many upsell opportunities (e.g. other production plants), additional related products, maintenance and other implementation and managed services and the churn rates are very low. The average order value will also grow over time due to increasing customer demand.

In the last decade, several companies specializing in securing operational assets have been founded. It’s an irreversible trend. The level of awareness for OT security is increasing and will continue to increase further as additional attacks happen and the damage impacts revenue for a range of manufacturing companies and other critical services such as electricity and water.

We’ve now seen one recent acquisition of an OT cybersecurity provider by an IT security firm and believe this trend will continue. As the market matures, and awareness and spending within the sector continue to increase rapidly, we expect to see more OT vendors and IT security firms acquiring the OT security leaders.

IT Security vs. OT security

There are different paradigms in IT security and securing industrial control systems: While the top priority in IT security is to protect the network and data, OT security vendors must ensure secure operations with no impact on the running operations. Unlike enterprise networks, industrial control systems often run for decades. It is not possible to take a traditional IT security product and use it to secure operational assets. Another challenge is to handle the complexity of very heterogeneous networks. A production site usually includes multiple automation vendors who all use proprietary communication protocols to monitor and control physical processes. The landscape typically differs in each industry vertical. Deep understanding of all processes as well as the communication protocols of sometimes twenty to thirty-year-old technology is required to succeed in this environment.

Most of the time the OT systems have a life-cycle of fifteen to twenty years and fixed structures make them inflexible and difficult to change. Since everything is operationally critical, system integrators often play a major role. While the ICS cybersecurity market is somewhat immature, the specialized products are maturing rapidly, and we believe that the company with the most active installations will learn quicker than the others.

Our Claroty Investment

Our investment in Claroty is a logical consequence of the fast-growing underserved market, its relevance to the automotive industry in general, and the market position Claroty has already established. Claroty emerged, in the past two years, as the distinct champion addressing the challenges of the OT security market in an unrivaled manner and with unprecedented market success. While the market is relatively young, Claroty has secured significant contracts and gained the trust of large multi-nationals in multiple vertical market segments from manufacturing to oil and gas, electric utilities, mining, pharmaceuticals water and more.

With headquarters in New York, and offices in Israel, Europe, and Asia, Claroty was founded in 2014 by CEO Amir Zilberstein, Chief Business Development Officer Galina Antova, and CTO Benny Porat. Amir has assembled an unmatched interdisciplinary team of security researchers with knowledge gained from both industrial careers and work within Unit 8200, an elite cybersecurity unit of the Israeli Defense Forces. Claroty emerged from Team8, Israel’s leading cybersecurity think tank and company creation platform, and is led by the kind of radically ambitious management team that we love to back.

Claroty’s growth is also impressive. The company expanded its bookings by more than 300 percent last year as demand for industrial cybersecurity solutions increased, and this followed a prior year of triple-digit growth. Based on this, Claroty can claim much of the credit for how quickly cybersecurity in OT networks has become established in production sites of the world’s largest and most influential companies.

For any of our investment decisions, but particularly in industrial sectors, a company’s partnerships are very important. Industrial customers are both cautious and cost sensitive and looking for a good cost-benefit (risk) ratio. They are looking for suppliers that have been tested, validated and approved by the major industrial automation vendors.  Customers are also more inclined to buy a solution package from companies bundling security into network infrastructure such as switches, or from companies that can bundle security capabilities with productivity features such as asset management.

Final Remarks

Given the challenges we have highlighted above, we believe an excellent partner network, deep knowledge in industrial automation and cybersecurity including broad coverage of ICS communication protocols, and the ability to execute are essential for a winner in this market. Sales cycles are rather long, but it’s a landgrab with high customer retention and upselling potential once an account is won. We are convinced the OT security market will grow rapidly over the two next years and Claroty is well positioned to address the challenges in this market. We are looking forward to working with the company to secure the safety and reliability of industrial control networks that run the world.

P.S: Special thanks to our intern Lenard Rüde for the help with the diligence and model.

ChargePoint raises $240 million to serve an anticipated flood of electric vehicles


Electric vehicle charging network ChargePoint has raised $240 million in a Series H funding round that attracted a diverse group of investors from the automotive, energy, financial, venture capital, utilities and even oil and gas industries.

New investors in the round include American Electric Power, Chevron Technology Ventures, Canada Pension Plan Investment Board, Daimler Trucks & Buses, GIC and Quantum Energy Partners. Existing investors include BMW i Ventures, Braemar Energy Ventures, Linse Capital and Siemens.

The latest fundraising effort comes during an aggressive growth period for the company, thanks to the growing number of automakers that plan to produce electric vehicles.

“The broader energy and mobility ecosystem has recognized that we are at a tipping point in the generational shift to transportation electrification,” ChargePoint president and CEO Pasquale Romano said in a statement.

ChargePoint, which designs, develops and manufactures hardware and software solutions across every electric vehicle use case, has raised more than $500 million to build out its network. Last year, the company’s Series G round of $125 million funded its expansion into Europe.

Today, ChargePoint’s network contains more than 57,000 independently owned public and semi-public charging spots.

Now, the company is preparing for an expected deluge of electric vehicles that will be introduced over the next four years. Automakers of every size and region, from GM and Volkswagen to Volvo and Jaguar Land Rover, have either launched an electric vehicle or are preparing to. VW, for instance, has said it has booked production for 15 million EVs.

ChargePoint is focused on meeting the needs of the everyday electric vehicle owner, as well as building out the infrastructure to maintain fleets. The company plans to use this injection of capital to expand its charging network in Europe and North America, expand options for fleets and improve the experience for EV drivers, the company said.

This means the company plans to build out features like integration with Amazon’s Alexa voice assistant or Tap to Charge, which lets drivers access information by using voice commands or start a charge from their phone.

RideCell expands funding round to $60 million


RideCell, a transportation software startup, has doubled its previously announced Series B funding round to $60 million, a sign that investors believe demand for cloud-based mobility platforms will grow as more companies try to scale up car-sharing, ride-hailing and even robotaxi businesses.

The company, which has developed a platform designed to help car-sharing, ride-sharing and autonomous technology companies manage their vehicles, announced it raised $28 million in May.

Activate Capital led this round; its co-founder and managing director Raj Atluru has joined RideCell’s board. Reinsurance group Munich Re’s ERGO fund, LG Technology Ventures, BNP Paribas, Sony Innovation Fund, Ally Ventures and Khosla Ventures joined this extended round. Denso also upped its investment in the Series B round.

Nearly half a dozen other companies had already invested in the Series B round, including Cox Automotive, Initialized Capital, Denso, Penske, Deutsche Bahn and Mitsui.

“Investor interest in cloud-based mobility platforms and autonomous vehicles increases almost daily as the disruptive potential of these new technologies are realized,” RideCell CEO Aarjav Trivedi said in a statement.

The company recently received a permit from the California Department of Motor Vehicles to test its Auro autonomous vehicles on public roads. RideCell acquired self-driving car company Auro in October 2017. Auro initially developed and operated driverless shuttles for private geo-fenced locations such as corporate and university campuses. The company has since expanded its focus to include passenger vehicle models and minivans, although it still plans to target low-speed urban use cases focused on solving last-mile transportation.

The company’s real-world trials will start on Ford Fusion vehicle platforms equipped with Auro’s autonomous driving system.

May Mobility expands to a third U.S. city


May Mobility  launched its first low-speed autonomous shuttle service in Detroit this summer. By March, the Ann Arbor, Michigan-based company will be operating in at least three U.S. cities.

The company, which just announced plans to expand to Columbus, Ohio, is planning to add another route in Grand Rapids, Michigan. It’s a rapid acceleration for a company that was founded less than two years ago.

May Mobility is different from other companies racing to deploy autonomous vehicles at a commercial scale. The startup, which was founded by veterans in the self-driving and automotive industry, has developed low-speed autonomous shuttles that are designed to run along a specific route in business districts or corporate and college campuses.

The company said it will bring four of its six-seat electric shuttles to Grand Rapids. The one-year pilot will begin March 2019.

This latest shuttle launch is part of a broader effort called the Grand Rapids Autonomous Mobility Initiative, a coalition of companies that includes Consumers Energy, French automotive supplier Faurecia, Gentex, Rockford Construction, Seamless and furniture maker Steelcase .

The aim of the program is to study how mobility impacts city infrastructure and prepare the community for autonomous vehicles. The program will also focus on how these autonomous vehicles improve or affect the mobility of elderly and disabled people.

The fleet will operate on a 3.2-mile section of an existing bus route that provides access to downtown and two of the city’s business districts. The route includes 22 stops, 30 traffic lights and 12 turns, including three left turns, according to the initiative.

Shuttles, which will be free for riders, will run complementary to the city’s existing DASH transportation fleet.

Fleet operations for the May Mobility vehicles will be housed at Rockford Construction’s
West Side offices within Circuit West, an area that boasts an innovative electric generation and distribution system.

May Mobility raised $11.5 million in seed funding in 2018 from BMW iVentures, Toyota AI and others. Trucks, Maven Venture and Tandem Ventures are also investors in the company.

Lunewave is pitching a new sensor offering better vision for autonomous vehicles

The investment arms of BMW and the Chinese search technology giant, Baidu,  along with a large original equipment manufacturer for the auto industry and a slew of technology investors, have all come together to back Lunewave, a startup developing new sensor technologies for autonomous vehicles.

The $5 million seed round, which the company just closed, will serve as a launching pad to get to market its novel radar technology, based on the concept of a Luneburg antenna.

First developed in the 1940s, Lunewave’s spin on the antenna technology involves leveraging 3D printing to create new architectures that enable more powerful antennas with greater range and accuracy than the sensing technologies currently on the market, according to the company’s chief executive John Xin.

Lunewave was co-founded by brothers John and Hao Xin and is based off of research that Hao had been conducting as a professor at the University of Arizona. Hao previously spent years working in the defense community for companies like Raytheon and Rockwell Scientific after graduating with a doctorate from the Massachusetts Institute of Technology in 2000.

Younger brother John took a more entrepreneurial approach, working in consulting and financial services for companies like PricewaterhouseCoopers and Liberty Mutual.

Lunewave represents the culmination of nine years of research the elder Xin spent at the University of Arizona applying 3D printing to boost the power of the Luneburg antenna. With so much intellectual firepower behind it, Hao was able to convince his younger brother to join him on the entrepreneurial journey.

“He has a strong desire to commercialize his inventions,” John said of his older brother. “He wants to see it in everyday life.”

Image courtesy of

Now the company has $5 million in new funding to take the technology that Hao has dedicated so much time and effort to develop and bring it to market. 

“With a single 3D printer in the laboratory version we can produce 100 per day,” John told me. “With an industrial printer you can print 1,000 per day.”

The first market for the company’s new technology will be autonomous vehicles — and more specifically autonomous cars.

Lunewave is focused on the eyes of the vehicle, says John. Currently, autonomous technologies rely on a few different sensing systems. There are LIDAR technologies, which use lasers to illuminate a target and measure the reflected pulses with a sensor; camera technologies which rely on — well — camera technologies; and radar, which uses electromagnetic waves to detect objects.

Startups developing and refining these technologies have raised hundreds of millions of dollars to tackle the autonomous vehicle market. In June, the camera sensing technology developer Light raised over $120 million from SoftBank. Meanwhile, LIDAR technology developers like Quanergy and Leddartech have raised $134 million and $117 million, respectively, and some studies have claimed that the market for LIDAR technologies was already a $5.2 billion last year alone.

Most companies working with autonomous cars these days use some combination of these technologies, but the existing products on the market have significant limitations, according to Lunewave’s chief executive.

John argues that the Lunewave technology can detect more objects in a wider field of view and at greater distances than existing products thanks to the unique properties of the Luneburg antenna.

Think of the antenna as a giant golf ball with a 360 field of “view” that can detect objects at greater distances than existing Lidar technologies because of the distance constraints on laser technologies.

Hao Xin with a Lunewave prototype

“LIDAR right now is at the end of the day because of its short wavelength. It does not function as well in poor weather conditions. Penetration of shorter wave lengths would be very difficult in poor weather conditions,” John said. “Our radar technology has the ability to function across all weather conditions. Our hardware architecture of our Lunenberg antenna has the best distance and the spherical nature of the device has the 360 detection capacity.”

The company came out with its minimum viable product in 2017 — the same year that it launched. It was one of the early companies in the UrbanX accelerator — a collaboration between Mini and — and is part of BMW’s startup garage program.

The company raised $5 million in two structures. Its seed financing was a $3.75 million equity round led by the automotive investment specialist McCombs Fraser with participation from Ekistic Ventures,, Plug and Play, Shanda Capital, Lighthouse Ventures, Baidu Ventures and BMW iVentures. But a portion of its capital came in the form of a $1.25 million non-dilutive government grant through the National Science Foundation . “In late 2016 that’s what helped us to jumpstart the company,” said John.

Now, the company just needs to fulfill Hao Xin’s dream of taking the product to market.

“We have the product,” John said. “It’s not just taking in money. Now it’s about [proof of concepts] and pilots.”

Online used car startup Shift raises $140 million

Shift Technologies, an online marketplace for used cars, has closed a Series D financing round of more than $140 million in equity and debt.

The round, which consists of about $70 million in debt and $71 million in equity, was led by automotive retailer Lithia Motors. Bryan DeBoer, CEO and president of Lithia, will join Shift’s board of directors.

Previous investors Alliance Ventures, BMW iVentures, DCM, DFJ, G2VP, Goldman Sachs Investment Partners and Highland Capital also participated. This new capital brings Shift’s total financing of equity and debt to $265 million.

Shift,  which is based in San Francisco, serves car buyers and sellers. The company, founded in 2013, has built a software platform that lets customers shop for cars, get financing and schedule test drives. Car owners can use the platform to sell their vehicle, as well. Shift says any car it buys must pass a “rigorous” 150+ point inspection.

The company plans to invest in its technology platform and scale its engineering staff from 35 to more than 80 people by the end of 2019, CEO George Arison noted to TechCrunch in an email.  Shift employs 380 people. The company’s platform has focused on scaling in California; it covers about 80 percent of that market. But the company has long had its sights set on expanding beyond the Golden State.

Shift is focused on, and is heavily investing in, its peer-to-peer business, in which the company acquires cars from individuals and then sells them. Buying, refurbishing and then selling cars online is a logistics-heavy business pursuit, and one that has seen a number of competitors come and go in the past several years. But Arison says the company has not just survived; it has grown. 

Shift didn’t provide revenue numbers. But Arison cited the company’s more than 70 percent revenue growth in the past six months as an example of the company’s success.

The company did have a partnership with rental giant Hertz,  but that has since ended. At the time, Shift was going to feature vehicles from Hertz’s fleet inventory. It was meant to be a win-win: Hertz gets access to a new retail sales channel and Shift benefits from the rental car company’s ready supply of lightly used cars.

The partnership ended after Hertz opened its own retail stores that competed against Shift

Trust but Vera-fy: Our Investment in Vera


There has been a surge in data breaches impacting many aspects of our lives, targeting tech startups and political parties to financial institutions owning some of the most critical consumer information. In a world where cloud is ascendent, supply chains are digitizing and integrating, and employees and partners expect to access data anywhere from any device, perimeter driven cybersecurity solutions are failing to provide sufficient protection. We believe there is a perfect storm for data protection for a few reasons.

First of all, there has been a massive shift in the infrastructure environment. According to Gartner, 90% of organizations will adopt hybrid cloud/on-premises infrastructure by 2020. Given cost advantages, efficiency and flexibility gains, the shift to cloud is a no-brainer, yet this also changes the way organizations need to think about security. Being inherently shareable, the cloud brings greater challenges with managing security and proper visibility.

Concurrently, more enterprises are adopting BYOD (Bring Your Own Device) policies to enhance employee satisfaction and reduce costs. 60% of organizations already allow employees to use their own devices at work, and an additional 14% plan to begin allowing within the next twelve months. According to research by European Commission, losing critical company data, especially customer data, represents the major concern for most companies that have not adopted BYOD.

As data is moving to the cloud, and employees are demanding access from their own devices, compliance and data privacy rules, such as GDPR, are increasing. GDPR, in particular, applies not only to Europeans but also any organization with business in the EU. GDPR effectively holds companies that store or transmit personally identifiable information accountable for the safekeeping of that information, and fines can total up to 4% of the annual global sales.

We are seeing additional trends in the manufacturing industry with the rise of powerful on-demand manufacturing platforms, such as Xometry, quoting parts instantaneously and helping companies get parts cheaper and faster. The workflows are getting digitized and supply chains are becoming more integrated and global. As a result, there is constant sharing of data among enterprises all around the world, no matter what sector they are in. As some of the data could contain valuable core IP in forms of CAD files, images, PDFs, among others, there is an increasing importance to protect, control access to and track data.

Given the challenges we’ve highlighted above, there is a need for a new approach that can protect the data at the file level, no matter where it travels. During our deep dive into the topic, we’ve had a chance to learn more about Vera. The team at Vera has built a powerful data-centric security tool encrypting individual files with specific and dynamic access controls that travel with the file across networks, ecosystems, and applications. Thanks to Vera, you can instantly revoke access to any document, take back control over sensitive emails, and lock down confidential data - even if it’s been copied, forwarded, downloaded or shared in platforms, such as Box and Dropbox. Vera can help secure most common enterprise file types, such as MS Office, PDF, creative content and CAD files seamlessly across Mac, Windows, iOS and Android.

Now is a pivotal point for Vera as Carlos Delatorre is joining as the CEO. Carlos brings more than 20 years of software industry leadership experience to Vera, and most recently served as Chief Revenue Officer at MongoDB, Inc., where he was a key member of the company’s IPO leadership team.

We are convinced that Vera stands out in the crowd thanks to its mature solution, solid customer base of blue chip companies like GE, Cisco and Capital One, and impressive management team who helped scale some of the most successful enterprise SaaS companies. We are thrilled to announce our investment in the Vera team and join other investors, including Battery Ventures, Sutter Hill and Hasso Plattner Ventures, to help shape the future of data security!

P.S: Special thanks to our rockstar interns, Brett Neustadt for the help in diligence, and to Benjamin Kleinschnitz in helping shape our initial thinking around data protection in manufacturing!

The Future of Mobility Just Got (Lime) Greener


Today, we’re thrilled to announce our investment in Lime and its two co-founders, Toby Sun and Brad Bao.

Here at BMW i Ventures, we spend a great deal of time thinking deeply about the future of mobility. What will the sidewalks and roads of tomorrow look like, and how will goods and people travel over them? We knew we’d seen the future when we met the team and experienced the product at Lime.

More than half of all urban trips are 3 miles or less. They’re the quick coffee runs we make in the morning; the last few blocks between our subway or bus stops and the office; the nearby lunch spots we visit in the afternoon.

These short, local trips define our daily travel, yet 72% of them are made in large, bulky vehicles that spend an overwhelming majority of their time parked or idling in stop-and-go traffic. We’d rather spend 10 minutes traveling 2 miles via Uber or Lyft than 20 minutes walking the same distance. What an inefficient use of space and fuel. Even the most efficient combustion or electric car will still consume 40-50x more energy per kilometer than an electric scooter.

Fortunately, the rise of electric bikes, scooters, and other personal vehicles threaten to upend this status quo, and we’re thrilled to help with our investment in Lime.

We see a future filled with cleaner air, less congested streets, and a more mobile population. Once cities and citizens realize how much cheaper, quicker, and more convenient micro-transportation is for everyone, whole lanes will be converted to e-mobility zones.

We have long seen a future where local urban trips will be made on electric, shared, on-demand vehicles, and we are confident that many of those trips will be on Lime e-bikes and scooters.

Maps, the crucial piece of autonomous puzzle, and our investment in Mapillary


High definition maps play a key role for autonomous driving, yet there are many challenges to tackle. Our investment in Mapillary, an independent provider of street-level imagery and map data, can help address these and make autonomous driving a reality.

Why are maps important?

World map by Gerard van Shagen, Amsterdam, 1689

World map by Gerard van Shagen, Amsterdam, 1689

Maps have played a key role for mobility and trade throughout the history. In ancient Mesopotamia, Babylonians engraved a map on clay tablets to find their way around the holy city of Nippur. Anaximander created a map helping Greek trade ships navigate their way through the Aegean sea. Yet, there is a flurry of recent news on maps... A coalition of carmakers (incl. our parent company), later joined by suppliers all around the world, acquired HERE for several billions dollars. Uber is investing $500M for its global mapping project. Google has acquired Waze for over a billion dollars and has been investing heavily through Waymo and Google Maps. Softbank led a $164M funding round for Mapbox very recently.

Maps, specifically high definition (HD) maps, are at the heart of the autonomous vehicle (AV) ecosystem for a few reasons. First of all, AVs need detailed maps for accurate and precise localization. Today’s GPS technology is far from being perfect, not to mention the issues with the “urban canyon” environment (Yes, this is why your uber shows up a few blocks away). This is especially true under challenging weather conditions, where driving without HD maps is compared to “putting a blind person behind the wheel” by industry analysts. Thanks to HD maps, AVs can match road furnitures and triangulate their position to get centimeter-level accuracy. Another important reason for having detailed maps is sensor redundancy. AVs are equipped with many sophisticated sensors (LiDAR, camera, radar, etc.) and some of these sensors can collect millions of laser points per second resulting in hundreds of GBs of data per hour. Maps can provide foresight, which can lower the workload on sensors and processors. Finally, maps can help AVs see what’s around the corner (e.g. latest accident or construction data) and further down the road. The sensors we’ve mentioned have a limited range- few hundred meters under even the best weather conditions. This means a handful number of seconds if you are driving 70 mph. Maps can integrate real-time traffic information to make the ride safer and more comfortable.   

On another note, maps are really interesting from an investing perspective, as map data could be created and consumed by AVs through a common cloud-based platform. This could create a virtuous cycle allowing the solution to get even better and more defensible over time with usage. As highlighted by another investor, “maps have network effects”.

What are the key challenges?

However, creating a detailed map with the precise location of every road furniture, incl. traffic signs, lane markings, among others, and updating it near real time (According to TomTom, about 15% of the roads change every year!) is not a trivial task and there are many key challenges. First, there is a need for huge deployment and we’d need millions of cars on the streets sending their sensor data back to the cloud to create a detailed and up-to-date map. Another challenge is the inconsistency of road furnitures among different locations, which makes the automation of the process even more complicated. The location and structure of the signs change even state by state, e.g. hanging traffic lights are very common in many states, whereas we have poles here, in California. This might sound simple, but think about the scale of this problem if you are a carmaker shipping vehicles all around the world. An additional challenge would be the process of stitching these images/videos from a variety of sensors. Even the angle of the sensors impact the process, not to mention inconsistent capturing quality of a variety of sensors on different vehicles.

On the other hand, today’s traditional map companies are operating in silos, by running their own survey vehicles around the world to collect data and updating their maps accordingly, which is another manual and cumbersome process. This approach is very costly as these survey vehicles are equipped with expensive sensors and they can only cover a small area as this requires a lot of effort. As a result, many different companies cover the same limited areas by incurring huge costs, and no single player ends up having the most accurate data. This results in very slow update cycles, and most of data installed in the vehicles becomes outdated quickly because of this. Additionally, there are cases where certain companies (municipalities, among others) are willing to share map data they’ve collected and can’t find a platform for that.

Our investment in Mapillary

Source: Mapillary AB

Source: Mapillary AB

As we were learning more about maps and their challenges, we’ve met Jan Erik Solem, the founder and CEO of Mapillary. Jan Erik has been active in computer vision space since the late 90s. His work on 3D reconstructions for his PhD turned into Polar Rose, a startup providing facial recognition solution running across mobile, cloud and desktop environments. Polar Rose was acquired by Apple in 2010 and the company’s technology helped power some of the latest face detection and recognition APIs and features in Apple products. Following the acquisition, Jan Erik ran a computer vision team at Apple before leaving to start Mapillary. Additionally, he has been a professor at Lund University, and written books about programming and computer vision (one of them still on the top lists after 5+ years).

Jan Erik started Mapillary late 2013 with Yubin Kuang (his former PhD student), Peter Neubauer, and Johan Gyllenspetz. Mapillary has an impressive team, including winners of the Marr Prize, one of the top honors for computer vision researchers. The team publishes their findings and openly shares their data and some of the code on Mapillary’s research website -

Jan Erik describes his vision at a blog post as follows:

“The core idea behind Mapillary is to combine people and organizations with very diverse motives and backgrounds into one solution and one collective photo repository, sharing in the open and helping each other. This means our awesome community, our partner companies and partner organizations, even our customers. That’s right, we’re incentivizing our customers to share their data into the same pool as everyone else, in the open, with an open license...

That we don’t have a mapping or navigation product means we can happily partner with any mapping or navigation company without being in competition. We’re neutral, open, and can work with anyone as long as it benefits our long term vision of visually mapping the planet. This means that photos and data from these partnerships will benefit OpenStreetMap too.”

Source: Mapillary AB

Source: Mapillary AB

Mapillary is a street-level imagery platform for generating map data at scale through collaboration and computer vision. Mapillary’s technology allows users to upload pictures in a device-agnostic way and generate map data anywhere. Unlike many other players, Mapillary doesn’t need expensive survey vehicles with special gear to achieve quality data. Mapillary has been investing the resources into developing computer vision algorithms, which are capable of of handling data from such a range of different sensors and compensate for any shortcomings of consumer-grade devices. Additionally, it is estimated that there will be 45 billion cameras by 2022, which could fuel the data capturing process further.

Mapillary has a strong network of contributors who want to make the world accessible to everyone by creating a virtual representation of the world. Anybody can join the community and collect street level photos by using simple devices like smartphones or action cameras. This means individuals as well as companies, governments, and NGOs. Recently, facilitating Microsoft’s upload of millions of images quickly, Mapillary’s technology helped with disaster recovery for hurricanes in Florida and Houston. Currently, Mapillary has more than 260M images uploaded, 4M km mapped, 190 countries covered and 22B+ objects recognized with computer vision.

Last May, the company released the Mapillary Vistas Dataset, the world’s largest street-level imagery dataset for teaching machines to see. The dataset includes 25K high resolution images, 100 object categories with global geographic reach under highly variable weather conditions. For autonomous driving, this means that cars will be able to better recognize their surroundings in different street scenes, which in turn helps improve safety. By using this training data and developing new approaches in computer vision research, Mapillary has the best results in semantic segmentation of street scenes, based on two renowned benchmarks.

We believe that there is a strong data network effect for Mapillary’s business. More engagement from community results in more engagement from customers, who contribute images, too. This is further catalyzed by improving computer vision algorithms, developed by Mapillary’s high-caliber research team. As a result, community members and customers extract more value in return and this drives further growth.

Given the challenges we’ve highlighted above, we believe just like Mapillary that there is a need for an independent provider of street-level imagery and map data, which could act as a sharing platform among different players as well. We believe that sharing data is crucial for accurate maps and safer autonomous vehicles as no single player has the necessary deployment in place to ensure this. We are convinced that any company should have access to most accurate map data, and the safety of AV passengers shouldn’t be a differentiator. With the world-class team, passionate community, and unique capabilities, Mapillary is well positioned to address the challenges and help make AV a reality. We couldn’t be more excited to lead this round of investment, and join Atomico, Sequoia and LDV, along with the new investors Navinfo and Samsung to shape the future of mapping!

Baris is an engineer with work experiences in venture capital and top-tier investment banking. At BMW i Ventures, Baris's investing scope encompasses a variety of areas, including industry 4.0, autonomous driving, mobility, AI, digital car/ cloud, customer digital life and energy services. Baris holds a Master of Engineering and Technology Management from Duke University along with an MBA (Dean's Fellow) from UNC Kenan-Flagler Business School, where he led VCIC, the world's largest venture capital competition. Please feel free to reach out on