Formula One Group is moving most of its infrastructure from on-premises data centers to AWS and standardizing on AWS machine-learning services—including Amazon SageMaker.
Using historical race data collected from cars over the past 65 years, Formula 1 data scientists are training deep-learning models that make race predictions and help teams optimize mid-race decisions. The models can predict when teams should pit their cars, determine the best timing for changing tires, and evaluate how drivers are performing.
Formula 1 then uses AWS data streaming, analytics, and media services to deliver insights about driver decisions and car performance to its more than 500 million fans.
Because Formula 1 runs its high-performance computing workloads in a scalable environment on AWS, the organization can innovate on the Formula 1 racing experience, car design, and more without worrying about capacity.
In 2014, Capital One had a fledgling private cloud capability and was also experimenting with AWS. In an effort to build the features its customers wanted as quickly as possible, the company chose to pursue AWS.
Security and compliance are critical to Capital One as a financial institution. To address these needs, Capital One built a cloud-risk framework and established a cloud-governance function before moving a single workload to AWS.
Capital One offers a cloud training program to engineers who work directly on AWS as well as non-technical stakeholders who need to be able to advocate for the power of cloud computing. The company now ranks in the top three of all enterprises globally for AWS certifications held by employees.
Using AWS, Capital One brings new products to market in weeks instead of months or years, uses machine learning to improve customer experience, and attracts top developers and engineers—all of which help the company achieve its most important goal: creating great experiences for its customers.
MLB has been collecting statistical data on its players and clubs for decades, and in 2015 it started using AWS to collect and distribute game-day stats to enhance the fan experience.
By using Amazon Sagemaker, MLB is empowering its developers and data scientists to quickly and easily build, train, and deploy machine-learning models at scale.
These models help MLB eliminate manual, time-intensive processes associated with recordkeeping and statistics, like scorekeeping, capturing game notes, and classifying pitches.
MLB plans to work with the Amazon ML Solutions Lab to continue improving Statcast—its tracking technology that analyzes player performance—including testing accuracy of pitch predictions and creating personalized viewer experiences.
MLB will continue to innovate using artificial intelligence. The organization plans to use Amazon Comprehend to build a language model that could create scripts for live games that simulate iconic announcers.
Matson built a flagship mobile application for global container tracking that allows customers to perform real-time tracking of their freight shipments. Other valuable features in the application include interactive vessel schedule searching, location-based port map lookups, and live gate-camera feeds.
All mobile devices access AWS via Amazon API Gateway. This provides highly available edge located endpoints for access into resources within Matson's existing virtual private clouds.
The AWS Lambda functions are designed using the microservices pattern and are modeled around specific ocean-based business contexts, such as shipment tracking and vessel schedules.
Amazon DynamoDB manages configuration as well as user-feedback configuration and user-feedback notifications sent from mobile devices. DynamoDB Streams provides real-time notifications to Matson's customer service team.
Matson can now offer customers an end-to-end serverless application to help track their shipments, and has no infrastructure to maintain.
The Roomba 900 series completes a cleaning mission in the home and returns to the dock for charging.
iRobot processes the home map, calculates the total floor space cleaned and the status code for the cleaning mission, and publishes the metadata to AWS IoT.
iRobot uses an AWS IoT rule to put the message into an Amazon Kinesis stream. From Kinesis, iRobot can process the cleaning mission data. Kinesis allows multiple teams to receive the stream of data.
AWS Lambda receives the cleaning mission metadata and parses the format to Amazon DynamoDB. Amazon Kinesis batches the mission data and stores it in Amazon S3. Amazon S3 is used as the iRobot data lake for analytics, where all message data is compressed and stored. Once the data is in Amazon S3, iRobot uses the AWS Analytics toolset. Amazon Athena allows iRobot to explore and discover patterns in the data without having to run compute resources all the time.
The cleaning mission is stored in Amazon DynamoDB and linked to a specific robot and consumer.
The consumer gets an alert that informs them of a successful Roomba 900 series cleaning mission.
BP's IT organization manages SAP applications used by thousands of employees worldwide for supply chain, procurement, finance, and more.
To improve speed and gain cost agility, BP used Amazon EC2 to migrate these core business apps to the cloud. In addition, the team built EC2 X1 instances to increase scale and to power their real-time analytics.
The team can now stand up systems on demand in hours instead of weeks or months. BP is seeing performance increases across the board, including a 40 percent speed improvement for the Lubricants ERP system.