Luxoft and Vantage Power Enable Powertrain Innovation and Safety with AWS IoT Telemetry

Constituents across the globe, from government groups to local advocacy organizations, insist that companies find ways to foster a low-carbon economy for the health and safety of the environment. Many vehicle and parts manufacturers in the transportation sector have committed to reducing harmful emissions from gas engines by investing in the electrification of vehicles.

Adoption of low-carbon technologies among the private and light-duty commercial sectors has increased over the years, promoting more development. By comparison, the heavy-duty vehicle sector has been slow to follow, causing a lag in innovation.

Governments are passing legislation to motivate change. London, for example, has committed to buying only zero-emission buses by 2025, upgrading their bus fleet to Euro VI standard by 2020, and ensuring all Ultra-Low Emission Zone vehicles are either hybrid or fully electric.

Manufacturers of heavy-duty vehicles and vehicle systems want to get new technologies to market fast in order to respond to the growing legislation and demand. This means that many Original Equipment Manufacturers (OEMs) might ship products earlier than usual, without the years of testing behind more established product streams.

This leaves OEMs and operators looking for a solution to track and monitor their new technology in the field after they’re deployed. By collecting operational data, they can prevent potential problems that might arise during the life of the vehicle. Tracking and monitoring allows high-quality servicing of parts, modelling of scenarios that can improve products, and proof of compliance.

Vantage Power wanted a vehicle telemetry system that would deeply integrate into the powertrain, vehicle control software, and other existing systems to collect rich operational data that could be used to tell a comprehensive story about how vehicles and parts were performing. They wanted to create a way to closely monitor their lithium-ion battery systems to help customers detect cell-level defects early and mitigate issues, potentially preventing the replacement of a lithium-ion battery that can cost over $50,000 per bus.  

Vantage Power worked with Luxoft, an AWS Partner Network Advanced Consulting partner and AWS IoT Competency partner, to create a telemetry platform called VPVision, which brings the AWS cloud platform to each connected vehicle. VPVision provides Vantage Power’s customers with a comprehensive, technical understanding of powertrain components including engines, batteries, control systems, electric generators, and motors. The system is fully customizable to specific engines and fleets.

Built around an Internet of Things (IoT) architecture, VPVision leverages AWS’ IoT services including AWS IoT Core, AWS IoT Greengrass ML Inference, and AWS IoT Analytics. These services are combined with Amazon Simple Storage Service (Amazon S3) and AWS Lambda to optimize processing of hundreds of thousands of data points per minute, 1hz frequency over cellular connectivity, to derive insights and predictive analytics models that can then be distributed to the vehicles to take preventative action in real time.  

Luxoft introduced AWS IoT Analytics and IoT Greengrass ML, enabling users to receive reporting and insights into over 6,000 data points from each vehicle. Adding this to the back-end architecture of VPVision allows Vantage Power to build and train machine learning models on AWS using historical real-world data while simultaneously testing it on live data streams from in-service vehicles. By connecting to AWS machine learning capabilities, Vantage Power can access a powerful environment for conceiving, testing, and launching new ideas faster than ever before.

“Together with Vantage Power and using AWS, Luxoft has developed a solution that can provide incredible insights and value using the data from connected vehicles,” says Sam Mantle, managing director of digital enterprise at Luxoft. “Within seconds, complex data is transformed into actionable information and shown on the platform. For instance, the sensors know when a hybrid vehicle is entering a low emissions zone and ensures its battery is fully charged ahead of time, extending the range and reducing pollution.”

“Battery-powered, heavy-duty vehicles are increasingly common, but successfully deploying and maintaining them at scale remains a challenge,” says Alex Schey, CEO of Vantage Power. “Using real-world data from the in-service batteries, we’ve developed a model that detects a failure mode in the cloud months earlier than we can today. We’re excited about how this platform can be applied to battery technology in all sectors to bring about a more robust and agile way of managing electric heavy-duty vehicles.”

Today, VPVision is offered as a standalone product with complimentary engineering services that enable heavy-duty vehicle manufacturers and sub-tier OEMs to integrate their own products.  

Lithium-ion batteries need to be balanced for optimal performance, which requires the battery to be at a certain charge level for a sufficient period of time. With VPVision, Vantage Power developed a machine learning model using Sagemaker Notebooks, part of AWS IoT Analytics, to automatically analyze a vehicle’s data and identify the ideal time and location for balancing. As the vehicle approaches its destination, the model changes the mode operation to ensure the ideal state of charge is reached by the time the vehicle is parked. This extends the operational life of the battery by an estimated 10%. This innovation was developed in the cloud and tested on live, real-world data, making sure it was optimized and validated before deployment.

Vantage Power uses VPVision to drive its business by speeding up the time to market for new innovations by over 6 months, providing safety and compliance controls using predictive analysis and reducing costs through remote diagnostics and edge computing.  

Using AWS IoT analytics to support VPVision’s back-end architecture ensures customers can collect and process the data they need to identify and address potential safety and compliance issues for individual vehicles.

Vantage Power and Luxoft developed a machine learning model that analyzed billions of data points collected from a fleet of double decker buses driven for over a year in real-world conditions. These buses use Vantage Power’s new hybrid powertrain with lithium-ion batteries for heavy-duty vehicles which consist of thousands of individual cells. If any of these cells suffer a fault known as excessive self-discharge, it reduces the stored charge, which can reduce the performance of the whole pack and affect vehicle operation. By understanding the known instances of the fault, the model can now detect high self-discharge one month earlier than the prior techniques allowed. When a fault did occur, it was diagnosed before attending the vehicle, which helped ensure an effective response with minimal downtime.  

Machine learning models built on AWS can be deployed to edge devices to reduce cloud transaction and compute costs. Vantage Power and Luxoft created a diagnostic model that analyses data on-vehicle and reports back when there’s an issue that needs attention. This minimizes the administrative burden traditionally experienced in monitoring, managing, scheduling, parts provisioning, fixing, and commissioning vehicles back into service faster. For example, if a refrigerated truck unexpectedly stops cooling, costs can be as high as $40,000 if the cargo temperature exceeds a threshold and the stock is scrapped. Costly breakdowns can be avoided by deploying predictive fault analysis to the edge and emerging issues are prevented from impacting operations.  

Learn more about IoT technology on AWS.