FROM THE CLOUD TO THE CAR: THE END TO END PICTURE · •Over-the-Air firmware/software updates is a...

Post on 14-Mar-2020

3 views 0 download

Transcript of FROM THE CLOUD TO THE CAR: THE END TO END PICTURE · •Over-the-Air firmware/software updates is a...

NXP and the NXP logo are trademarks of NXP B.V. All other product or service names are the property of their respective owners. © 2017 NXP B.V.PUBLIC

SEPTEMBER 24TH, 2019

DR. RAJEEV ROY, NXPMICHAEL JOHNSTON, NXP

FROM THE CLOUD TO THE CAR:THE END TO END PICTURE

PUBLIC 1

AGENDA• Market trends• Use cases• Network evolution and landscape• Security considerations• Vehicle telematics• Summary

PUBLIC 2

Automotive Industry Megatrends

PUBLIC 3

38

V E H I C L E B I G D ATAo p p o r t u n i t i e s

MILLION

Represents global dataSources: *Strategy Analytics, 2019; **ABI Research, 2018

Shipped in 2018* Vehicle data generated per hour**

TERABYTES4+ VEHICLE DATACONNECTED VEHICLES

Connected vehicle penetration*

50%2019

2025

40% 73%2018 2025

PUBLIC 4

Vehicle Data Opportunities Will Transform the Automotive Industry

New Revenue StreamsUp to $750B* for data-driven services by 203077.4% millennials** willing to pay for updates

Enhanced Safety and SecurityFault detection & notificationIntrusion detection and preventionCrash detection / emergency response

Improved User ExperiencesPersonalization, comfort and conveniencePost-sale feature upgradesLocation-based services

Reduced CostsPredictive maintenanceReduced warranty / recall exposureFleet management

Sources: McKinsey & Company, Monetizing Car Data, 2016; IHS Markit forecast, 2018

PUBLIC 5

Use cases

PUBLIC 6

Over-The-Air (OTA) use case

• Over-the-Air firmware/software updates is a key trend in the industry

• Trend to move OTA Management function in Gateway ECU− Centralized management of OTA deployment in-vehicle− Interface to OEM servers − Security is paramount

• Utilizing OTA mechanism to deploy new features via SW in field (Agile SW deployment)− Build performance overhead into hardware− In-field, test & deploy new customer features as use cases

emerge

PUBLIC 7

Automotive SDN use case • Central configuration of data-plane

switches in IVN−Support both TSN and BE traffic−Support switching and routing

• Use cases:−Dynamic network changes E.g. Adding new services via OTA while

maintaining optimum network efficiency −Error handling E.g. Fast failover

• NETCONF CNC example:

Data path

PUBLIC 8

Analytics Typical use cases (1)

• Predictive maintenance− Identify vehicle maintenance issues before they

occur − Schedule maintenance when needed− Quickly identify the root cause of any problem− Machine Learning (ML) at the edge to detect

safety relevant maintenance issues

• Intrusion Detection and Prevention (IDPS)− Uses anomaly detection techniques to identify malicious

activity or faults on vehicle networks− Detects static anomalies (non ML)− Contextual anomaly – ML based− ML at the edge to detect attacks in real time

PUBLIC 9

• Crash detection− Used to detect if a crash occurs− Indicates the severity of crash− Indicates likelihood of serious injury− Capture moments before crash− ML at the edge to detect crash real time

• Usage based insurance− Behaviour policy pricing− Premium based on driver use of vehicle− ML determine driver premium depending on

several risk factors− ML at the edge to detect driver and usage

Analytics Typical use cases (2)

PUBLIC 10

Predictive DataAnalytics

Analytics: Handling Data• On-Vehicle processing

− Capture: Data to service (raw data to speed, temp, etc) Data logging

− Process: Predictive data analytics (limited)

• Deviation detection • Signal correlation, comparison to reference modules

Aggregate Diagnostic Trouble Codes (DTC) with other data and time stamps

− Share: Compression before offload

• Off-Vehicle Cloud processing− Correlation of data across multiple vehicles− Advanced predictions using historical and warranty

data− Part defect history

SOCCloud

NetworkServices

On board data streams

On board data streams (sensor data)

Repair history

Aggregated vehicle data

Model data

Model data (limited)

Off-VehicleOn-Vehicle

Predictive Data Analytics

ShareProcessCapture

Hardware support to analyse network

streams

Heavy math vs Decision Tree

Optimize available

bandwidth

PUBLIC 11

Network evolution and landscape

PUBLIC 12

Ethernet Ecosystem

Courtesy: Ethernet AllianceNote: Original image (link) is modified to have less text

The vehicle as a part of the larger Ethernet Ecosystem

PUBLIC 13

From LAN to WAN to IVN

Access Aggregation/Edge Metro/Core-Edge

Residential

Enterprise

Mobile

xDSL ATM

PDH

PDH/Frame Relay

IP/MPLS

ATM

SDH

TransportPDH

PUBLIC 14

From LAN to WAN to IVN

Access Aggregation/Edge Metro/Core-Edge

Residential

Enterprise

Mobile

xDSL

Ethernet

IP/MPLS

ATM

SDH

Transport

Ethernet

Ethernet

PUBLIC 15

From LAN to WAN to IVN

Access Aggregation/Edge Metro/Core-Edge

Residential

Enterprise

Mobile

xDSL

IP/MPLS

Ethernet

SDH

Transport

Ethernet

Ethernet

Ethernet

PUBLIC 16

From LAN to WAN to IVN

Access Aggregation/Edge Metro/Core-Edge

Residential

Enterprise

Mobile

IP/MPLS

Ethernet

OTN

Transport

Ethernet

Ethernet

EthernetProvider Backbone

PUBLIC 17

From LAN to WAN to IVN

Access Aggregation/Edge Metro/Core-Edge

Residential

Enterprise

Mobile

IP/MPLS

Ethernet

OTN

Transport

Ethernet

Ethernet

EthernetProvider Backbone

WiFi

Wired

Mobile

PUBLIC 18

Day in the life of a packet – A typical infra view

OEM Data Center Access Edge/Provider Backbone Edge/Provider Access Mobile RadioAccess

Vehicle Wireless Gateway

Application

TCP/UDP

IP

C-ETH C-ETH

S-ETH

B-ETHMPLS

OTN

B-ETH

IP

C-ETH

S-ETH

IP IPGTP/PDCP

UDP/IPRLC

IPPDCP

UDP/IPRLC

Application

TCP/UDP

IP

Target ECU

Copper Copper Copper CopperOptical CopperRadio

TCP: Transmission Control ProtocolUDP: User Datagram ProtocolIP: Internet ProtocolC/S/B-ETH: Customer/Service/Backbone EthernetMPLS: Multi Protocol Label SwitchingOTN: Optical Transport NetworkGTP: GPRS Tunneling ProtocolPDCP: Packet Data Convergence ProtocolRLC: Radio Link Control

PUBLIC 19

Wireless Connectivity Landscape

Ultra-short Range

Short Range

Wide Range

CAT-M1 NB-IoT5G

PUBLIC 20

Evolution of the Cellular Base Station

PUBLIC 21

Security considerations

PUBLIC 22

Secure Comms

WAN0

WAN1LAN0

Telematics Control Unit

(TCU) Gateway

LAN0

Subnet 192.168.4.x

Subnet 192.168.3.x

Subnet 192.168.2.x

88.11.10.11

78.14.10.19

OEM Server

VLAN1

VLAN2

VLAN3

VLAN4

APP

TLS

TCP

IP

Ethernet

IP

Ethernet

IPsec

APP

TLS

TCP

IP

Ethernet

IPsec

MACSEC MACSEC

IPsec and TLS can both be end to end

PUBLIC 23

Vehicle telematics

PUBLIC 24

Internal vehicle communications

Domain controller architecture Zonal architecture

PUBLIC 25

Example ECU architectures

Secure Ethernet Gateway

Telematics Control Unit (TCU)

PUBLIC 26

Summary

PUBLIC 27

Summary

• Market trends− Data is king!− Data volume, access and handling are key defining points

• Use cases− Enablement of new features is not a luxury, it’s something the user expects− OTA, SDN, remote diagnostics and analytics are key drivers

• Network evolution and landscape− Transport of a bit of data from the source to destination involves a myriad of technologies− 5G allows new levels of flexibility and scalability to enable new use cases− Ethernet is the common denominator

• Security considerations− A portfolio of techniques is available to protect data− End to end and/or point to point

• Vehicle telematics− Handling the data within the scope of the vehicle itself is non-trivial− TCU and Gateway ECUs are two of the key elements which complete this picture

PUBLIC 28© 2019 Cloudera, Inc. All rights reserved.

28

NXP-CLOUDERA VEHICLE EDGE2AI ANALYTICS & MACHINE LEARNINGArchitecture for Unlocking the Value of Vehicle Data

ANALYZE7 •Self-Service Business Intelligence (BI)

•Fleet Analytics

LEARN8

• Historical vehicle data• Historical maintenance records• Historical usage characteristics• Historical failures

Model Inputs

9 DEPLOY

ENTERPRISE TRANSACTION DATA

Design, MFG, Dealer Service, Warranty, etc.

5

ENRICH

Connected Car 1 Connected Car NREAL-TIME

ACTION

ACT1 VEHICLE EDGE ANALYTICS

TRANSMIT2

ENTERPRISE DATA LAKE

3

4

6

CDF

D A T A I N M O T I O N

V E H I C L E S E R V I C E -

O R I E N T E DG A T E W A Y

STORE, ENRICH & PROCESS

CDH D A T A A T R E S T

C l o u d e r a D a t a S c i e n c e

W o r k b e n c h

LEARN

ANALYZE

PUBLIC 29

Thank you

NXP and the NXP logo are trademarks of NXP B.V. All other product or service names are the property of their respective owners. © 2017 NXP B.V.