hello my name is John turdiv and I'm a Senior Solutions architect at AWS in this video we'll discuss AWS iot fleet-wise and how it can help to collect and analyze vehicle data in a standardized way we will Begin by highlighting the trends in the industry and some of the challenges in dealing with vehicle data next we will explain what AWS iot fleetwise is how it works and present its capabilities we will also describe where AWS iot fleetwise fits in a connected vehicle platform on AWS we will continue with the discussion of AWS iot fleetwise use cases and how many of our customers are using this service to address complex real-world challenges finally we will showcase AWS iot fleetwise using an EV battery health monitoring example and share useful resources to get you started using AWS iot fleetwise a number of factors have made vehicle data workloads Big Data workloads vehicle data is exploding exponentially accelerating the need for more intelligent methods that determine what data signals are important enough to collect and send to the cloud first there are increasing numbers of connected vehicles McKinsey forecast that by 2030 95 percent of new vehicles sold globally will be connected second each of those Vehicles will have increasing numbers of sensors generating data and each of those sensors will have capacity to generate richer data in the quantity of terabytes of data per hour finally with the industry Evolution to our software-defined vehicles data becomes not only useful but also critical component for an ever-expanding list of features and Digital Services for customers here are some of the specific challenges our customers have with collecting vehicle data data fragmentation fragmentation and silos are prevalent in the automotive industry due in part to how vehicles are made OEM sourcing Hardware software and connectivity systems from hundreds of suppliers there are thousands of proprietary signals for in vehicle communication for example signals like a GPS location or vehicle's speed can take many different names formats and types making it nearly impossible to create insights across a fleet of different vehicle models and brands for example 65 mile per hour in Seattle is 104.6 kilometer per hour in Vancouver as complex vehicle sensors like EV batteries cameras and lidars become standard in most Vehicles data fragmentation is likely to become an even bigger challenge as every OEM will need to collect data to train AIML models per sensor and per vehicle type a very expensive and time-consuming proposition data tsunami you might have heard about the data tsunami coming from connected devices and vehicles are no exception vehicles are transitioning to higher levels of autonomy over the next decade L3 autonomous vehicles AVS will have dozens of AIML models looking after your comfort safety and autonomy these models require a tectonic shift towards collecting camera radar and lidar data from production Vehicles not just test vehicles automakers need a solution to look for high value low redundancy AV data from millions of cars that will be used for continual AIML model training and Improvement additionally transferring all of this data to the cloud continually is cost prohibitive across a fleet of production vehicles with each vehicle generating terabytes of data per hour data delays while new vehicles are still under warranty automakers are responsible for the entire cost of repairing problems but do not have access to the data they need to predictively diagnose fleet-wise vehicle quality issues for systematic issues caused by faulty design or manufacturing it takes a long time to get enough data from customers reproduce the issue in a lab and escalate the fixed manufacturing by that time the issue might already have broad customer impact resulting in Costa recalls legal fees and broken trust with customers Plus with the Advent of semi and full AVS automakers need a solution that allows for near real-time fleet-wide anomaly detection that will allow for quick intervention AWS iot fleetwise is a fully managed service that enables automakers and Fleet operators to standardize collect and transfer vehicle data to the cloud at scale it's comprised of two components an open source software called The iot fleetwise Edge agent and iot fleetwise Cloud where you can create Cloud resources to model select and collect vehicle data you can think of iot Fleet wise as a data management layer that abstracts all the underlying device management and connectivity so that you can focus on your data-driven business cases instead benefits of AWS iot fleetwise include analyze standardized fleet-wide vehicle data access the unique data format of any vehicle structure it and make it understandable without having to develop a custom system developers build virtual vehicle models in the cloud and apply a common data format to structure and label vehicle attributes sensors and signals with an aggregated data structure automakers can access fleet-wide insights and identify Trends while using their cloud-based applications and services nearly all vehicles today use a standard in-vehicle Communications protocol called controller area network bus or can bus for short each vehicle model has a unique and proprietary canvas configuration which automakers can import to the AWS Management console AWS iot fleetwise reads the unique can bus data signals from each vehicle then applies a standard data format across the fleet improve data relevance with intelligent data collection use intelligent data collection to transfer only high value data signals to the cloud this helps keep costs low and gives access to more useful data with AWS iot fleetwise automakers can reduce the amount of data transferred to the cloud from each vehicle while not losing any signal quality within the AWS Management console automakers select which data to transfer such as data from safety equipment or any other data generated by the vehicle sensors then they Define rules and events for when to transfer data based on parameters such as weather location or vehicle type detect and mitigate problems more quickly by surfacing Vehicle Health Data in near real time surface selected data about widespread vehicle quality issues and take corrective action quickly such as notifying the manufacturing group to help mitigate further spread by detecting problems early on Automotive companies can reduce service and warranty costs help avoid large recalls and maintain customer Trust within the AWS Management console automakers can select what data to transfer to the cloud based on certain conditions such as notifying the L3 AV operations group about every autonomous disengagement event in near real time here's the flow of how AWS iot fleetwise works 1. build virtual representations of your vehicles in the cloud two connect the vehicle to the cloud using the edge agent which facilitates Communications between vehicles and the cloud three Define what data to collect and under what conditions using campaigns campaign contains data collection schemes you define a campaign in the cloud and deploy to a vehicle or Fleet campaigns give the edge agent software instructions on how to select collect and transfer data to the cloud or the edge agent collects and transfers the data to the cloud making it available to analytics tools here's an example architecture for connecting vehicles to AWS iot Fleet wise let's review its components one develop and install the edge agent for AWS iot fleetwise based on a reference implementation The Edge agent allows users to test simulated vehicle Data before integration or runs as an application to connect remotely to Athleta vehicles 2. create a semantic digital twin of the vehicle in AWS iot fleetwise by defining a vehicle model consisting of vehicle attributes such as model year and engine type standardizing vehicle data format and defining relationships between signals in AWS iot fleetwise provides a foundational vehicle data structure for creating data collection campaigns three create campaigns with condition based or time-based data collection schemes AWS iot fleetwise deploys active campaigns to Target vehicles to acquire sensor data from the vehicle network based on a defined data collection schemes or the edge agent applies inspection rules to upload vehicle data back to the database iot fleetwise data plane through AWS iot core a fully managed service that connects iot devices to the cloud the data plane persists the collected data in Amazon timestream or Amazon S3 for further analysis and five analyze Trends and patterns to generate actionable insights with AWS analytics services including Amazon quicksite for business intelligence Amazon managed grafana for data visualization Amazon Athena for interactive queries and AWS glue for data integration you can also build ml models using Amazon sagemaker let's discuss the features of AWS iot fleetwise in more depth beginning with the edge agent The Edge agent software running in vehicles facilitates communication between vehicles and the cloud while vehicles are connected to the cloud The Edge agent software continually receives data collection schemes and collects data accordingly The Edge agent provides C plus plus libraries that enable you to run the application on your vehicle it is compatible with most Linux platforms and qnix support is available from icas The Edge agent has a small footprint of about 5 megabytes and is provided on GitHub under the Apache License version 2.0 technical details such as porting and getting started is available in the edge agent developer guide it is also possible to deploy The Edge agent on AWS ec2 instances to simulate your connected vehicles we have numerous AWS partners that build Automotive solutions to optimize vehicle connectivity safety and autonomy by working with an AWS partner you can take advantage of solutions to streamline your iot projects reduce the risk of your efforts and accelerate time to value these AWS Partners include nxp and Renaissance instructions are provided for running AWS iot fleetwise Edge agent on an nxp s32g vnp rdb2 development board or Renaissance rcar S4 spider board and deploying a campaign to collect OBD data icast offers Edge agent support for platforms such as qnix neutrino one of the most widely used operating systems in the automotive environment let's explore AWS iot fleetwise campaign feature next campaigns allow you to Define schemes to transfer only high value vehicle data to the cloud you can Define condition-based schemes to control what data to collect such as in-vehicle temperature values that are greater than 40 degrees you can also Define time-based schemes to control how often to collect data features of campaigns include fully managed data collection across fleets of vehicles or a single vehicle fast and reactive ability to apply the policy and start retrieving the data set within minutes flexible semantics for defining fleets using rules based on vehicle attributes like make model location manufacturing date fuel type body type and so on flexible semantics for defining data collection policy when to collect what to collect and for how long to collect AWS iot fleetwise vehicle modeling feature enables you to build virtual representations of your vehicles and apply a common format to organize vehicle signals AWS iot fleetwise supports the connected Vehicle Systems alliances vehicle signal specification that you can use to standardized vehicle signals signals signal catalogs vehicle models and decoder manifests are the core components that you work with to model your vehicles let's review these key Concepts signals are fundamental structures that you define to contain vehicle data and its metadata a signal can be an attribute a branch a sensor or an actuator for example you can create a sensor to receive in-vehicle temperature values and store its metadata including sensor name data type and a unit a signal catalog contains a collection of signals signals in a signal catalog can be used to model vehicles that use different protocols and data formats vehicle models or model manifests are declarative structures that you can use to standardize the format of your vehicles and the Define relationship between signals in the vehicles vehicle models enforce consistent information across multiple vehicles of the same type you add signals to create vehicle models decoder manifests contain decoding information for each signal in vehicle models sensors and actuators in vehicles transmit low-level messages in binary data with decoder manifest AWS iot fleetwise is able to transform binary data into human readable values here we see an example vehicle model that consists of several attributes such as serial number type capacity and signals such as voltage current charging station and so on this example campaign defines the types of signals to collect under various conditions for instance voltage current and temperature are collected every 30 seconds additional detail is collected when the battery temperature is below 30 or higher than 90. furthermore All Battery faults are collected when issues occur this campaign helps us gain insights into the state of the EV battery health such as anomaly detection failure prediction and many others as stated earlier timely insights such as these allow us to identify vehicle quality issues and take corrective action quickly such as notifying the manufacturing group to help mitigate further spread by detecting problems early on Automotive companies can reduce service and warranty costs help avoid large recalls and maintain customer Trust let's talk about data store options available in AWS iot fleetwise Amazon timestream and Amazon S3 at launch in September 2022 AWS iot fleetwise provided Amazon timestream as a data persistence mechanism which is a Time series database primarily built to demonstrate and analyze how data changes over time providing the ability to identify Trends and patterns in near real time Amazon timestream provides a new real-time use cases which can give for example Fleet operators a holistic view of their Telemetry data via campaign deployed by AWS iot fleetwise Amazon timestream offers the following benefits no servers to manage or instance a supervision software patches indexes and database optimizations are handled automatically capable of ingesting trillions of events daily the Adaptive SQL query engine provides a rapid point in time queries with its in-memory store and fast analytical queries through its magnetic store built-in analytics using standard SQL with added interpolation and smoothing functions to identify Trends patterns and anomalies and all data is encrypted in flight and addressed using AWS Key Management Service with customer managed keys Automotive companies are searching for more efficient ways to simplify data collection from the vehicles Amazon S3 support for AWS iot fleetwise helps optimize the cost of data storage and also provide additional mechanisms to use vehicle data within a performant data Lake data processing pipelines visualization dashboards and other improvements to Downstream data services Amazon S3 offers highly performant and durable data management capabilities which help with unlocking new Revenue opportunities from fleets building machine learning data sets and creating predictive maintenance models to detect and resolve problems in near real time Automotive companies can use these new capabilities to gain insights on things like driving behaviors infotainment interactions and long-term maintenance needs for electric vehicle fleets Amazon S3 object storage for AWS iot fleetwise supports two industry standards data formats for Big Data implementations these are Apache parque and JavaScript object notation or Json Json is a standard human readable text-based format for representing structured data using JavaScript object syntax customers can use this format when they need to maintain relational data in the payload though there is slight storage and compute overhead to implementing this format most data Engineers will use Apache parquet format for vehicular Telemetry data as it's an open source flexible and scalable format offering efficient data storage and retrieval the format is suitable for data compression and encoding schemes in a variety of common programming languages with Amazon S3 customers can unlock online analytical processing or all app capabilities through batch data analysis with multi-dimensional data points this helps to continuously improve using historical data from across fleets of vehicles creating differentiation for the operator implementing predictive maintenance in their Fleet data engines can Implement tool sets using their common data processing to extract transform and load the data into an automotive data Lake from several different sources of data providing a centralized all-app storage for data scientists this flexibility allows data Engineers to bring vehicle data directly into other AWS services like Amazon Athena and AWS clue which provide abundant opportunities to enhance and enrich the Telemetry data using services like Amazon Athena and AWS glue also allows for formatting this data for use within machine learning models for example customers can continuously improve their predictive maintenance models range estimates or energy-based routing for Ev batteries Based on data stored in Amazon S3 from a battery monitoring system or BMS Amazon S3 helps customers secure data from unauthorized access with encryption features and access management tools S3 encrypts all object uploads to all buckets S3 is the only object storage service that allows you to block public access to all of your objects at the bucket or at the account level with S3 block Public Access S3 maintains compliance programs such as PCI DSS HIPAA high-tech fedramp EU data production directive and fisma to help you meet regulatory requirements AWS also supports numerous auditing capabilities to monitor access requests to your S3 resources AWS provides many services and reference architectures for creating an end-to-end solution for our Automotive customers this is right from the production to an end-of-life processing AWS identifies six main elements of a connected vehicle platform as shown here let's see how AWS iot fleetwise Works within a connected vehicle platform on AWS first the vehicle is manufactured and you can onboard vehicles to the platform using AWS iot core second for connectivity to the cloud and control you can use AWS iot core and AWS iot Greengrass third since different vehicle Brands models and components you generate data in different formats you will need AWS iot fleetwise for a data abstraction layer this will allow for Universal data interpretation once the data is in the cloud and easier fleetwise insights fourth in the cloud you will need services like AWS iot device management to help manage the fleet using features like Fleet Hub or dashboards and jobs for OTA updates fifth also in the cloud this is where you can build applications for business differentiation using tools like Amazon quicksite and Amazon for grafana and sixth spanning vehicles to Cloud you have security and privacy with services like AWS iot Defender that keeps communication secure here is how a connected vehicle platform looks from an architectural perspective starting at the bottom you have your device layer where you manage functions such as Hardware abstraction OS device onboarding and management going up a level you have your connectivity layer where you create a secure connection from vehicle to Cloud using protocols like mqtt in between the vehicle and the cloud you have both software abstraction and data abstraction these layers allow you to get a unified view of software and data the operations layer in the cloud is where you monitor deploy and perform customer support the applications layer is where you build applications that differentiate your business and your product AWS iot fleetwise allows automakers Tier 1 suppliers telematics solution providers and other Auto providers to collect vehicle data transform it and then transfer it to the cloud to gain insights about their fleets of vehicles there are endless possibilities for what we can do in the cloud with data transferred by AWS iot Fleet twice here are some of the most relevant use cases for automotive companies today use analytics and machine learning to improve models for autonomous driving and Advanced Driver assistance systems analyze Vehicle Health to quickly identify potential maintenance issues monitor EV battery health make in-vehicle infotainment systems smarter and improve Fleet operations and driver safety our customers are using AWS iot fleetwise service to address complex real-world challenges here are some highlights of such cases automakers are embarking on a digital transformation journey to become more agile efficient and innovative as part of this transformation Continental created Continental Automotive Edge a modular multi-tenant hardware and software framework that connects the vehicle to the cloud Continental collaborated with Amazon web services to develop and scale this framework Continental needed a way to collect Fleet data to optimize cicd processes supported by the Continental Automotive Edge the company is working with AWS to integrate AWS iot fleetwise with Continental Automotive Edge and introduce data-driven CI CD to help validate software deployments for both test and production vehicles Continental is expecting to save up to 50 percent of field operation test costs by collecting data for critical use cases for customer fleets this reduces the cost of regular software updates by up to 25 percent and delivers software updates faster to Continentals customers LG Electronics is a leading Global provider of vehicle components and solutions LG is providing a vehicle data platform and Mobility Services using AWS iot fleetwise LG is developing Mobility Services and upgrading solutions to help automakers and Mobility service providers deliver new customer experiences Edge to Cloud Solutions allow LG to rapidly develop Mobility Services to provide vehicle Data Insights and understanding to develop Mobility Services and upgraded Solutions LG needed to develop Edge to Cloud solutions that could rapidly collect vehicle data including a vehicle Big Data platform that provides data analysis and Mobility Services by adopting AWS iot fleetwise LG is able to intelligently collect OBD can data in near real time and simulate vehicle data for virtual Fleet status using AWS data pipeline LG has also developed a vehicle data platform using real and simulated data the benefits of this solution are as follows collect data quickly regardless of card type based on the industry standard VSS data structure rapid prototyping and cost-effective validation of Mobility Services that can increase in size and understanding of vehicle behaviors now let's take a hypothetical example of a customer that wants to use iot fleetwise to collect EV data now that you understand how customers get started with iot Fleet wise picture yourself as someone that manages an electric vehicle Fleet and wants to monitor the health of the battery subsystem across thousand vehicles in your production Fleet detect any issues inspect root cause create a fix and validate the fix We Begin this demo by showing you how iot fleetwise standardizes EV data signals while also protecting proprietary information and Technology let's begin by creating a data model that represents several signals you can collect from your EV Fleet we call this signal catalog in iot fleetwise and it uses the connected Vehicle Systems alliances vehicle signal specification standard format available on GitHub as you can see we have created several vehicle signals extending what's currently available from the VSS standard for example we create a new way to represent voltage and temperature per battery module that's an area of values in addition across all the modules we introduce signals to capture min max and other aggregate values our goal is to create something generic to all or most EV architectures out there similar to what the OBD format does for internal combustion vehicles then you can create a subset of the signal catalog called the model manifest to model a specific vehicle type or a specific vehicle subsystem from here you can create vehicles that represent this model manifest next as you can see in the vehicles tab we have instantiated thousand vehicles that belong to the EV production Fleet which includes Vehicles connected using Linux based Hardware now that you're done with data modeling you can start to create and deploy data collection campaigns let's first deploy an event-based campaign that detects unhealthy Vehicles across our entire EV Fleet by uploading data only when the BMS ECU indicates it has an active DTC or the state of health drops below 75 percent you start your day at the office by visiting the production Health dashboard from this dashboard you are immediately able to tell the number of vehicles marked as unhealthy for your data collection campaign logic the dashboard provides a unique ID of the unhealthy Vehicles using this high-level Insight you assign these three unhealthy vehicles to a fleet called the inspection Fleet you can do that manually or programmatically based on your scale and complexity involved then create a deep dive data collection campaign that you would deploy specifically to unhealthy vehicles this data collection campaign is a time-based campaign that enables you to collect EV data at high frequency to help inspect what is causing your vehicles to be in unhealthy state you look at the rich set of signals you define in your time-based campaign and immediately notice a few anomalies you can see that the state of health is at 70 percent and there are a few modules exhibiting high temperatures at about 45 degrees Celsius you also notice that the outdoor temperature is 42 degrees which indicates the vehicle is being driven in a hot environment finally you notice that the cooling fan in the BMS was not enabled in the midst of these modules failing due to high temperature you take these insights to your ECU software development team for scrutiny and find out that there was a bug in the ECU configuration that led to the cooling system not engaging properly the development team creates a new ECU configuration and wants to validate the fix in a test environment finally create a validation data collection campaign that you deploy to your test Fleet of vehicles to validate the fix in the ECU configuration the team deploys the ECU configuration fix to simulated Vehicles as well as to a test of fleet vehicles placing them in hot environment to reproduce the issue and using a test dashboard like this the team is able to validate the ECU turning on the cooling system as expected when the outdoor temperature starts to exceed 40 degrees celsius allowing the modules to retain normal operating temperature in our demo we assume that the fleet has only one EV model that we can map to existing VSS attributes what if we had a fleet of multiple EV models types or even brand names collecting data from such an EV Fleet would be a very hard problem to solve as each automaker has a proprietary way of reporting battery and charging status data we believe that there's a need to create new standards that evolve the OBD to electric vehicle data there are some useful resources to get you started with AWS iot Fleet wise AWS developer guide offers quick start and in-depth explanation of AWS iot fleetwise GitHub repository contains source code and detailed documentation for AWS iot fleetwise Edge agent you can sign into your AWS Management console to begin modeling your vehicle and defining data collection campaigns reach out to your AWS account team who can arrange a discussion with our iot specialists I hope you found this video informative thank you and until next time