Proceedings

Data, Analytics, and the Transformation of the Automotive Industry

Paul Ballew, Vice President and Chief Data and Analytics Officer, Ford Motor Company

Data and analytics are transforming most industries, but one of the biggest transformations is taking place in the automotive industry. There are substantial pressures affecting all facets, both internal and external, to Ford, which are requiring data and analytics to address these pressures. Paul Ballew, vice president and chief data and analytics officer for Ford Motor Company, will discuss the elements as well as Ford’s strategy to leverage data and analytics to transform the business and develop future products and services.

Deploying Controls for a Self-Driving Taxi

Alan Mond, Senior Mechatronics Engineer, Voyage

Voyage deploys self-driving taxis in senior living communities. Their autonomous taxi fleet is currently offered to seniors daily in private communities in California. This presentation will highlight the methods, Simulink^®, and more importantly, failures and workarounds to deploy a functioning control stack quickly. By leveraging open source software, Simulink, docker containers, and fast iterations with Embedded Coder^®, Voyage was able to bring a service that people can safely use in their daily routine.

Vertical AUTOSAR System Development at John Deere

Jim Sachs, System Engineering Manager, Enterprise Electronics Group, John Deere

Increasing electrical and electronic complexity coupled with the growing need for greater automation and connectivity are driving the need for a highly integrated system development process that enables agile software development. This presentation will share John Deere’s approach of a highly vertically integrated process and tool chain that support agile development, model-based software design (MBSD), AUTOSAR integration, and system engineering.

Automated LiDAR Point-Cloud Annotation for Sensor Verification

Nathan Kurtz, Autoliv and Arvind Jayaraman, MathWorks

In the world of automated driving, sensing accuracy is of the utmost importance, and proving that your sensors can do the job is serious business. This is where ground truth labeling has an important role in Autoliv’s validation process. Currently, annotating ground truth data is a tedious manual effort, involving finding the important events of interest and using the human eye to determine objects from LiDAR point-cloud images. This talk presents a tool that was developed in MATLAB^® to alleviate some of the pains associated with labeling point-cloud data from a LiDAR sensor and the advantages that tool provides to the labeler. The capabilities of the tool are discussed, including assisting users in visualizing, navigating, and annotating objects in point-cloud data; tracking these objects through time over multiple frames; and then using the labeled data for developing machine learning based classifiers. The talk also describes how the output of the labeling process is used to train deep neural nets to provide a fully automated way to produce vehicle objects of interest which can be used to find false-negative events. To do this with a human analyst takes as much time as to play back the entire data set. However, with a fully automated approach, it can be run on many computers to reduce the analysis time. This presentation shows the time savings as well as the accuracy of the labels achieved and how this approach provides substantial benefit to Autoliv’s validation process.

Model-Based Calibration Optimization Using Machine Learning Algorithms

Shehan Haputhanthri and Shuzhen Liu, Ford Motor Company

Calibration optimization of any production vehicle requires hardware prototypes, which could cost up to millions of dollars to be built, demand a lot of engineering time, and add a substantial cost to the vehicle powertrain (PT) design validation (DV) process. Electrified powertrains with their sophisticated supervisory control strategies and thousands of tunable calibration parameters are particularly challenging and time consuming to calibrate. High-fidelity computer models with embedded production software code could be used for initial calibration efforts to reduce the number of prototypes and engineering time required for powertrain calibration. This presentation explains how MATLAB^® and Simulink^® vehicle models, machine learning algorithms, MATLAB Distributed Computing Server™ capabilities, and high-performance computing were used for model-based calibration optimization. Over 10,000 different designs worth of 4000–6000 hours’ worth of simulations was completed in less than 15 hours to optimize fuel economy and other attributes.

AUTOSAR Based Software Development Workflow Using a Model-Based Approach

Raghavendra Manjunatheshwara, Embedded Software Engineer, Halla Mechatronics

This presentation introduces an AUTOSAR based software development workflow using a model-based approach. The workflow covers the following aspects:

• Importing the AUTOSAR software component description file (.arxml) into MATLAB^®
• Creating the Simulink^® model skeleton
• Developing the algorithm using Simulink^® and Stateflow^® as well as generating code using Embedded Coder^®
• Generating the software unit design document using Simulink Report Generator™ and conducting static analysis using Polyspace Bug Finder™
• Conducting model-in-the-loop (MIL) and software-in-the-loop (SIL) testing using a verification and validation toolbox

This optimized AUTOSAR based workflow can enhance software deployment in the automotive industry.

Rapid Engine Control Prototyping Using Simulink Real-Time and Speedgoat Target Hardware

Roopak Ingole, Manager of Advanced Embedded Software, Advanced Dynamic Systems and Controls – R&D, Cummins, Inc.

Cummins, Inc. has complementary business units that design, engineer, manufacture, distribute, and service engines and related technologies. Their R&D relies on tools that enable engineers to innovate on new products and technologies and bring them to market faster than other players. They have adopted Model-Based Design using desktop simulation and code generation products from MathWorks. Apart from simulation, another key factor for faster research is rapid prototyping of controls algorithms. To replace decade-old prototyping tools, Cummins evaluated several off-theshelf tools and selected Simulink Real-Time™ and Speedgoat hardware. This has enabled the company to develop new algorithms, evaluate alternative sensors and actuators, and validate them in test cell environments. Cummins and MathWorks collaborated to leverage existing Cummins production hardware and ECU, bypassing support of Simulink Real-Time. Cummins is expanding the usage of this toolchain for various use cases, including hardware-in-the-loop (HIL) simulation.

What’s New in Automated Driving System Toolbox

Mark Corless, Automated Driving Segment Manager, MathWorks

In this presentation, you will learn how MATLAB^® and Simulink^® provide a development environment for components in advanced driver assistance systems (ADAS) and automated driving (AD) applications. You will see examples that you can use to get started developing:

• Vision detection algorithms with deep learning
• Sensor fusion algorithms with recorded and live data
• Longitudinal (ACC) and lateral (LKA) control algorithms with synthetic sensor data

Design and Test Traffic Jam Assist, A Case Study

Seo-Wook Park, Principal Application Engineer, MathWorks

Traffic jam assist systems require a combination of longitudinal control, stop and go management, and lateral control with lane following control. This presentation will show you how to:

• Design a MPC-based lane following and longitudinal controller
• Specify driving scenarios using the Driving Scenario Designer app
• Synthesize sensor detection using a vision and radar sensor model
• Design a sensor fusion algorithm
• Run tests in simulation

Demystifying Deep Learning: A Simplified Approach in MATLAB

Brett Shoelson, Principal Application Engineer, MathWorks

In this overview, Brett Shoelson will demonstrate MATLAB^® features that simplify the complex tasks required to implement deep learning solutions without the need for low-level programming. In doing so, he’ll explore concepts of deep learning by building and training neural networks to recognize and classify objects, as well as to figure out the drivable area in a city environment.

Along the way, you’ll see MATLAB features that make it easy to:

• Manage extremely large sets of images
• Import and use pretrained models like Alexnet or GoogLeNe
• Perform classification and pixel-level semantic segmentation on images
• Automate manual effort required to label ground truth

Tackling Fleet Test Data with MATLAB

Will Wilson, Senior Application Engineer, MathWorks

Can your data analytics technology keep up with the rising data intake from a connected test fleet? Are you able to find interesting events in stored data, and zoom in and out with ease?

In this talk, Will Wilson will demonstrate how to implement engineering applications quickly and efficiently with MATLAB^® to: 

• Automatically detect events of interest and zoom in for signal-level insight
• Verify analytics on both the desktop and cluster
• Deploy the analytics to keep up with the continuous intake of test data

Predictive Maintenance: Approaches for Estimating Remaining Useful Life

Alec Stothert, Engineering Manager, MathWorks

Creating accurate estimates of RUL (remaining useful life) is essential to successful predictive maintenance deployments.  However, statistical uncertainty around RUL and a myriad of potential algorithms that could be used result in a complex design space. In this talk, Alex Stothert will show common approaches for estimating RUL and validating RUL models. He will also introduce the new Predictive Maintenance Toolbox™ including pre-built reference examples you can use to get started.

Integrating MATLAB Analytics into Enterprise Applications

Arvind Hosagrahara, Principal Engineer, MathWorks

As the size and variety of engineering data has grown, so has the capability to access, process, and analyze those (big) engineering data sets in MATLAB^®. With the rise of streaming data technologies, the volume and velocity of this data has increased significantly, and this has motivated new approaches to handle data-in-motion. We’ll discuss the use of MATLAB as a data analytics platform with best-in-class frameworks and infrastructure to express MATLAB based workflows that enable decision making in “real-time” through the application of machine learning models. We’ll also demonstrate how to use MATLAB Production Server™ to deploy these models on streams of data from Apache^® Kafka^®. The demonstration shows a full workflow from the development of a machine learning model in MATLAB to deploying it to work with a real-world sized problem running on the cloud.

Five More Cool Things You Can Do With Powertrain Blockset

Mike Sasena, Product Manager, MathWorks

In last year's conference, this presentation discussed five ways you could use Powertrain Blockset™ to accelerate your powertrain systems and controls development. This year, Mike Sasena will explore additional use cases, including:

• Automating engine model parameterization
• Braking controls development
• Aftertreatment system testing
• Battery cooling circuit testing
• Vehicle dynamics modeling

Virtual Engine Calibration: DPF Regeneration Example

Peter Maloney, Senior Principal Engineer, MathWorks

To reduce calibration workload resulting from legislative and market requirements, MathWorks now provides controller and plant models, in addition to calibration optimization tools. MathWorks has also developed automated processes to convert measured data into calibration parameters to achieve accurate vehicle-level powertrain simulation models. To illustrate the integration of controller models, powertrain plant models, and automatic calibration optimization tools, the virtual calibration of a multiple-injection engine plant model and controller state estimator will be demonstrated. The plant and estimator models will then be used within a larger vehicle drive-cycle simulation model to predict the fuel economy impact of pass-through fuel injections used for DPF regeneration events.

Electrified Powertrain Design Exploration

Kevin Oshiro, Senior Application Engineer, MathWorks

Designing electrified powertrains is a challenging task that includes evaluating, analyzing, and comparing different powertrain architectures. In this presentation, a workflow will be demonstrated on how to systematically build, design, and optimize powertrain plant models and control strategies in order to facilitate architecture selection. The following topics will be discussed:

• Using the Powertrain Blockset and Simscape™ to efficiently build system-level powertrain architectures which enable the process for Model-Based Design
• Develop supervisory control strategies that are robust over varying driving conditions, enhance performance, and are real-time implementable
• Utilize optimization techniques to simultaneously optimize both component and control parameters as a system

A P2 HEV architecture will be used as an example during this presentation.

Accurate Simulation of EV/HEV Power Electronics Switching Events for HIL Testing

Joel Van Sickel, MathWorks

Learn how to use FPGA technology to perform real-time simulation of FEM-based motor models while capturing power electronics switching events. This presentation describes a system-level model of motor and inverter deployed to run at a rate of 1 MHz on an FPGA using HDL Coder™ combined with Simulink Real-Time™ and Speedgoat real-time hardware.

Data, Analytics, and the Transformation of the Automotive Industry

Paul Ballew, Vice President and Chief Data and Analytics Officer, Ford Motor Company

Data and analytics are transforming most industries, but one of the biggest transformations is taking place in the automotive industry. There are substantial pressures affecting all facets, both internal and external, to Ford, which are requiring data and analytics to address these pressures. Paul Ballew, vice president and chief data and analytics officer for Ford Motor Company, will discuss the elements as well as Ford’s strategy to leverage data and analytics to transform the business and develop future products and services.