Securing the connected vehicle through standards

• Introduce myself & W3C
• Explain W3C Auto standards
• Data protection policy
• Extend more broadly to layered protection

W3C

• Standards body for the Web
• Founded by inventor of Web - Tim Berners-Lee
• MIT, Beihang, Keio, ERCIM
• 450+ members including Baidu, Alibaba, Huawei, Tencent and Xiaomi

Why is W3C doing Auto Standards?

• Need for a robust application ecosystem
• Future of transportation will require interoperability
• Reduce fragmentation by competing proprietary approaches
• More developers for the Web than any other platform
• Web has proven itself and transformed industries
• Connected vehicles will be relying on Web services
• W3C has a long history of providing higher level interfaces to developers

Rich application ecosystem

• Telematics (car data) - W3C VISS (Vehicle Information Service Specification)
• Notifications
• Media
• CDN/Caching
• LBS/Nav
• Payments
• other areas under consideration, awaiting critical mass

W3C Vehicle Information Service Specification

• Service on the vehicle
• Common data model (JSON)
• ~1000 signals, more to be added, extensible
• Web Sockets, next gen also HTTP REST
• granular accesss control per application
• Requests or Subscribe (stream)

Implementers

• KDDI/ACCESS
• Mitsubishi Electric
• Volvo
• Bosch
• Renesas
• Visteon
• Samsung/OCF
• Others?

Data at the Center of it all

• Electrification - payments, range, capacity, charge rate
• Usage instead of individual ownership models
• Autonomous Vehicles - Artificial Intelligence data driven decisions
• Insurance carriers, maintenance, fleet management, other industries and government regulators
• V2v, V2i, V2x - SmartCities
• Delivering content, products and services digitally

Data Task Force formed

• Explore various off-boarding challenges to enable marketplace
• Handling privacy concerns (GDPR) with Policy language
• Increase power of data with Semantic Web Technology
  BMW/Eurecom ontology work on our common data model
• Increase value of data - metadata on sampling methodology (edge computing), accuracy, availability

Layered Security

Thus what is of supreme importance in war is to attack the enemy's strategy. - Sun Tzu

• Application
• Data (service)
• Network
• Host OS

Data Security

• Personal Identifying Information
• Interactions with underlying vehicle (W3C VISS, access control)
• Need to control app information access
• Policy Language
  who gave access to what information to share with whom for what purpose?

Application Manifest Policies

• control data access to specific signals
• govern what information is off-boarded
• what other vehicle services can be accessed (location, media, payments)
• what computing resources (libraries, files, etc) can be used
• which internet services can application talk to
• ...

Application Manifest Policy Enforcement Examples

• data access - confined subset per app
• off-board information - Web Application Firewall
• vehicle service access - AppArmor, SELinux etc
• computing resources - also AppArmor, SELinux etc
• internet services - local DNS zone file excerpts, Public TLS Cert, Firewall
• monitor, log and report any deviations from expected norm
• disable suspect applications automatically
• Varying solutions using same policies a major advantage

Advantages to common/standards based approach

• Already explained common data model and access methods enable code portability across platforms
• Lower integration costs
• Wider peer review
• Best practices shared across industry
• AutoISAC or similar incident reporting model

Follow Up

• Chinese automanufacturer, supplier, provider involvement in W3C Auto
• Sponsor research at MIT
• Collaborate on EV adoption
• Contact Ted Guild or Xiaoqian Wu