BMW Delivers IoT Services to 1M Car Owners by Using IBM Cloud Foundry

The vehicle manufacturer has created a secure data sharing platform based on Cloud Foundry and IBM Watson, exploring more scenarios for digital innovation.
Why read this?
Use Case for Cloud Foundry:

An automotive manufacturer transforms digitally by providing connected apps and value-added IoT services to drivers of BMW vehicles.

Business or Technical Result:

Launched in May 2017, the BMW CarData app has opened a new connectivity model to the company—secure "data sharing as a service"—relying on IBM Cloud Foundry (Bluemix). Today, there are 40+ data-driven apps being developed and tested at the BMW ConnectedDrive marketplace through CarData, offering third-party digital services (e.g., real-time traffic monitoring). Over 1 million car owners are already using these apps, securely sharing their telematics data with automotive software providers and organizations such as insurance companies.

Lessons learned:

Rather than building a custom app for each customer need, setting up a marketplace—where third parties provide services instead—was a better long-term decision.

What else is in the stack?

IBM Cloud Foundry (a.k.a. IBM Bluemix), IBM Watson

Company Description:

BMW is a German manufacturer with more than 100-year history of producing automobiles and motorcycles. As of 2018, the company has over 134,600 employees worldwide and a reported revenue of €97.48 billion. In 2018 alone, the company delivered almost 2.5 million vehicles; 142,600+ of them were "electrified."

Cool fact about the company:

BMW produced aircraft engines until 1945. Luxury brands Rolls-Royce and Mini are subsidiaries of the company.

Too much car data

As automobile manufacturers embrace the Industry 4.0 trend, newly produced vehicles are being equipped with an increasing amount of sensors. These devices then collect all sorts of data—such as location, mileage, or the health and status of individual car components.

According to McKinsey, a connected car generates around 25 GB of data per hour. To make use of the information being collected through car sensors, most manufacturers develop apps and services that cater to various driver needs—from remote vehicle monitoring to cognitive driver assistance.

Car sensors generate more data than media streaming services (Image credit)

As one of the top automotive manufacturers, BMW is investing heavily into digital transformation. For instance, sensors implemented within BMW’s vehicles monitor telematics across numerous parameters:

  • 63 condition data points (fuel and fluid levels, GPS location, mileage, battery voltage, etc.)
  • 14 usage data points (trip distance, charging status, average fuel consumption, etc.)
  • 2 event-based data points (emergency and breakdown situations)

Some of this data is available to drivers through the BMW Connect app, which provides destination routing, time-to-leave notifications, remote control and monitoring, etc.

An overview of the BMW Connected app (Image credit)

Since delivering its first “connected app” back in 1998 (BMW Assist), the German automobile giant recognized the difficulties in having to create an app for each of the scenarios and use cases a driver may need. Though developing a broad set of connected services by the manufacturer itself could be an option, it may not be sustainable due to long delivery cycles and other reasons.

“BMW is a huge company, we don’t move superfast. (But) if you are an innovative fast company, you are going to be the first.” —Christian Clauss, Project Manager, BMW

So, in 2014, the company launched ConnectedDrive Store, which serves as a marketplace providing approved third-party apps for various driver needs. Hovewer, one of the questions on the agenda was the sharing of the teledata described above to third-party developers in a secure, trusted way. The company needed a unified, easy-to-use solution that will enable transparency for drivers and comply with existing regulations (such as GDPR).

 

An open data platform on IBM Cloud Foundry

In 2016, the manufacturer partnered with IBM to explore the cognitive capabilities of IBM Watson (available through the Bluemix platform) for BMW i8 sport cars. However, the partnership resulted in a completely different project.

On May 30, 2017, BMW unveiled CarData, an add-in for ConnectedDrive, introducing an open data platform for drivers and third-party developers. CarData relies on IBM Cloud Foundry and Watson, serving as a “neutral server” and enabing drivers to manage and control telematics data collected from their BMW vehicles. Car owners can then use the app to securely share their vehicle information to third-party service providers of choice.

How the data is shared through BMW’s data center (Image credit)

Within this system, BMW enables an integrated interface where third parties can provide value-added services to car owners. With the integration of IBM Cloud Foundry (Bluemix), third-party developers have access to both BMW CarData APIs and IBM’s services catalog. This way, the addition of IBM Watson enables the creation of apps with advanced analytics or predictive maintenance based on vehicle wear and tear.

The platform opens the door for unique innovative services. Typical service providers, such as car insurers, can develop a dynamic app that automatically uses and verifies gas mileage, driver behavior, and overall vehicle health / usage. Secure sharing of this data can save time, as well as may cost less than traditional operations.

During his presentation at TU-Automotive Europe, Christian Clauss of BMW explained how location-based services could benefit from CarData. In his example, he proposed a mobile car wash provider, which could be accessible via the BMW ConnectedDrive portal.

“This company would need a GPS position of a parked car. The service to a BMW driver would be exterior cleaning (at the place) where the car is actually parked. As a result, the BMW driver can now use a new service that wasn’t possible before.” —Christian Clauss, BMW

Christian Clauss of BMW at TU-Automotive Europe 2017 (Image credit)

Telematics data related to overall car health could be useful to vehicle maintenance providers, too. With BMW CarData, they can develop apps for scheduling and alerting drivers about service appointments. Potentially, the CarData platform can improve the connectivity of 10+ million BMW vehicles—those that are equipped with built-in SIM cards.

As the solution was aimed to comply with EU’s GDPR regulations, car owners can withdraw their consent at any time. In addition, according to Christian, a BMW vehicle can be registered to a single owner. In case ownership changes, the data of the previous owner is deleted once a new user connects to the vehicle with CarData.

“If you sell your car or if a new user connects to the car, then everything will be terminated. There is a one-button delete-everything functionality.” —Christian Clauss, BMW

According to BMW CarData FAQ, telematics data is stored “for a limited period and then deleted, once the statutorily specified period has expired.” CarData also provides a report of the data shared and an archive that contains a list of all the telematics data collected for this driver. The use of the CarData app is free of charge.

 

Kubernetes-Training-Courses

Results: 40 apps in 45 countries

BMW CarData was initially launched only in Germany back in May 2017. Then, the coverage has expanded to Austria, France, Luxembourg, and the Netherlands. As of May 2019, ConnectedDrive Store is available for 45 countries, including the United States.

According to Christian, more than 60 third-party vendors have registered with BMW CarData and, among those, 40 data-driven apps are being developed and tested. Since the launch, BMW CarData has already enabled over 1 million car owners to consume IoT services based on their securely shared vehicle data sets.

“Protecting vehicle data is part of our understanding of ‘premium’ in a highly connected vehicle.” —Peter Schwarzenbauer, BMW

Peter Schwarzenbauer of BMW at DLD Conference 2017 (Image credit)

There are also plans for additional functionality, such as collecting time-series data and analysis of telematics information over a period of time. The company foresees the concept CarData to enter the areas of infotainment and smart homes.

The scenarios CarData is involved require sharp attention to the safety of sensitive data, as well as privacy, making it a case for considering blockchain as a tool to use. The same relates to any other teledata flows within BMW’s apps and systems.

So, it was very natural that on May 2, 2018, BMW became one of the founding members of the Mobility Open Blockchain Initiative (MOBI). With this announcement, BMW, as well as other automotive manufacturers and technology partners, will be exploring the use of blockchain to address various needs in the automotive industry, such as data sharing, supply-chain tracking, autonomous payments, etc.

 

An intelligent personal assistant

Above, we mentioned that BMW has been working with IBM since 2016 in an effort to develop an intelligent personal assistant for the automotive manufacturer’s i8 hybrid sports cars. In this partnership, BMW wanted to rely on IBM Watson’s cognitive and machine learning capabilities (available via Bluemix) to understand a driver’s habits and preferences.

As IBM put it, “Watson will learn the i8 owner’s manual and with the natural language capabilities to be able to understand the driver’s questions, he’ll understand what you’re asking and provide answers in a conversational style.” This collaboration was going hand in hand with a report released by IBM noting that 54% of consumers wanted their vehicles to be able to automate and personalize their experience.

On the other hand, later, in September 2018, it was revealed that BMW’s intelligent personal assistant is being built on top of Microsoft’s Azure stack. According to BMW’s blog post which appeared earlier this month, “Microsoft’s Virtual Assistant Solution Accelerator built on Azure provides the necessary technological basis, leveraging Microsoft Azure cloud and AI capabilities, such as Bot Framework and Cognitive Services.”

The intelligent personal assistant on Microsoft Azure (Image credit)

According to BMW, the AI tool is available for cars equipped with the BMW Operating System 7.0, starting from March 2019. For instance, it can be found (and was pioneered) inside BMW 3 Series.

So, it is a bit unclear now whether BMW will continue exploring how IBM Watson can be used as a cognitive assistant for the i8 hybrid sports cars only—or the Azure-based assistant will be standard across all the fleet models. Nevertheless, looks like the Watson collaboration is still ongoing. In March 2018, IBM announced Watson Assistant targeted at “automakers and hotels, and electronics and other companies.”

“You’re going to be talking to BMW or your Whirlpool appliance or whoever implements this for you.” —Bret Greenstein, IBM

 

Other related digital initiatives

Besides CarData and the intelligent assistant, BMW is also collaborating with several other members of the Cloud Foundry Foundation. In 2016, the car manufacturer partnered with Intel and Mobileye to develop fully autonomous driving vehicles. (Intel acquired Mobileye in 2017.) There were 40 prototype vehicles in the test fleet running in 2017. An additional hundred was going to be added at a later date.

The partnership has already reached SAE Level 2, where partial automation is achieved through sensors and cameras in and around the vehicle. SAE Level 4, where high automation is made possible through not only sensors and cameras, but also real-time data mapping, is expected by 2021. SAE Level 5, which is fully automated driving without the need of driver intervention is estimated to be available by 2030.

BMW has also been working closely with Microsoft for years. The most recent example is the Open Manufacturing Platform—a joint project announced in April 2019. Created with Microsoft Azure IoT, it aims at improving logistics and supply chain at a “smart” factory.

The Open Manufacturing Platform, a high-level view (Image credit)

BMW is also using data collected from its vehicles for predictive maintenance using tools from Pivotal, a founding Platinum member of the Cloud Foundry Foundation. “We run many sensors on our cars simultaneously. They feed the data to forward the so-called ‘memories’ with diagnostic travel codes,” emphasized Dirk Rueger of BMW during EMC World 2015.

According to him, BMW strives at finding the right correlation patterns between the ‘forwarded memories’ to predict malfunctions that may occur with cars and their consequences. The primary goal of this initiative is to detect malfunctions of car components at the earliest stage possible.

BMW makes use of the Pivotal Big Data Suite to handle and analyze the large amounts of data coming from the vehicles produced. Additionally, data analytics enables BMW to manage its business more proactively and accurately. “There’s a challenge in synchronizing different development cycles,” he explained. “On the one hand, we have car development, and, on the other, we have IT development, which is very rapid.”

“Traditionally, car development is usually measured in years and IT in months. That means the cycle of digital development is up to 12x faster than car development. The challenge right now is integrating digital services with late car development cycles.” —Dirk Rueger, BMW

According to Dirk, BMW has also created an augmented reality app that enables drivers to take photos of car controls and receive information about the items. “We use computer vision algorithms in order to detect the control and route directly to the corresponding item in the user manual,” noted Dirk. “The content of the user manual is almost simultaneously shown.”

So, as we can see, the amount and diversity of BMW’s initiatives that can be labeled as digital transformation is indeed impressive. Without doubt, this would not have been possible without ecosystem collaboration.

 

Want details? Watch the videos!

Table of contents
  1. What is BMW CarData? (1:15)
  2. How does BMW CarData provide a B2B service? (2:50)
  3. How do third parties take part in BMW CarData? (4:20)
  4. BMW CarData demo (7:14)
  5. The results since launching (9:50)
  6. What’s next? (11:40)
  7. Q&A: How does BMW CarData data sharing work? (13:50)
  8. Q&A: How much will it cost third parties to use the B2B service? (14:25)
  9. Q&A: How will BMW CarData interact with temporary vehicle owners? (15:31)

 

Further reading

 

About the experts

Christian Clauss is ConnectedDrive Project Manager at BMW. He has over 10 years with the German automobile manufacturer. During this time, Christian has brought many features to markets worldwide, allowing the customers to seamlessly integrate their daily routines to their connected cars. Currently, he leads the development and roll out of BMW CarData.

 

Peter Schwarzenbauer is member of the board of management at BMW. He has over 30 years of international experience with premium automotive brands. In the past, Peter has served as President and CEO of Porsche Cars North America Inc. and as a member of the board of management for Audi AG. In 1984, he graduated from the University of Applied Sciences in Munich with a degree in Business Administration.

 

Kai Barbehön is Vice President of Electric/ Electronics Central Control Units, Wire Harness, Power Supply at BMW. He initially joined BMW in 1998. Over the last 20 years, Kai has served the automotive manufacturer as the Section Manager for Engine Management Systems, Director of Electric/ Electronics, Manager of Platform Technologies, and Manager of Process Chain Driver Environment Experience and Electric/ Electronic. Prior to BMW, he worked as a Software Engineer for BOSCH.

 

Dirk Rueger is Head of Aftersales Business Analysis at BMW. He initially joined BMW in 2011 as Head of Integration and Quality Assurance Aftersales. Dirk is a graduate of the Technical University of Berlin.

 
 


This blog post was written by Alex Khizhniak, Carlo Gutierrez, and Sophia Turol.
  •  
  • 26
  •