AWS CEO Any Jassie made a number of new announcements related to AI and ML services. AWS has had a machine learning service for a while now and provides AI building blocks that customers can leverage to build solutions. The AWS AI/ML service ecosystem can be categorized into two groups: core infrastructure services and managed services. AWS has put a lot more focus on the managed services recently and re:invent 2020 announcements reflect this.
The AI/ML use case are still emerging but we believe these services will be integrated into the functionality of virtually all applications in the future. AWS has also made a number of ML algorithms available as a service built using commodity data. The organizations can use their proprietary data with these services to differentiate themselves and develop unique intelligent applications.
Core Infrastructure Services
The new chipset Trainium is the second customer machine learning (ML) chipset announced by AT after the launch of Inferentia in 2019. Trainium is meant to provide the best price-performance for training ML models in the cloud. The chip shares the same AWS Neuron SDK as AWS Inferentia.
In addition to delivering the most cost-effective ML training, Trainium will offer the greatest performance with the most teraflops of computing power for machine learning in the cloud. Also, the chip is optimized for various deep learning training workloads for applications, such as image classification, translation, voice recognition, natural language processing, among others.
Managed Machine Learning Services
SageMaker Data Wrangler
This new AWS service is designed to speed up data preparation for machine learning and AI applications. According to its developers, it reduces the time it takes to aggregate and prepares data for machine learning (ML) from weeks to minutes. The service contains over 300 built-in data transformations so that one can quickly normalize, transform, as well as combine features without having to write any code. Customers can use this service to import and inspect data to identify the various types, recommend transformations, and apply it to the entire data set.
SageMaker Feature Store
Amazon SageMaker Feature Store provides a purpose-built repository that makes it much simpler to name, organize, and find and share machine learning (ML) features with teams. It provides a unified store for features during training and real-time inference without the need to write additional code or create manual processes to keep features consistent. Amazon SageMaker Feature Store integrates with Amazon SageMaker Pipelines to create, add feature search and discovery to, and reuse automated machine learning workflows. As a result, it’s easy to add feature search, discovery, and reuse to your ML workflow.
Amazon SageMaker Pipelines
Amazon SageMaker Pipelines is the first purpose-built (CI/CD) service for machine learning (ML) that customers can leverage to create, automate, and manage end-to-end ML workflows at scale. It can automate different steps of the ML workflow, including data loading, data transformation, training and tuning, and deployment. With SageMaker Pipelines, different teams within the organization can share and re-use workflows to re-create and optimize ML models. The logging capabilities of this service can also help with compliance as it logs every step of your workflow, creating an audit trail of model components such as training data, platform configurations, model parameters, and learning gradients.
Amazon DevOps Guru
Amazon DevOps Guru is a Machine Learning (ML) powered fully managed service that makes it easy to improve an application’s operational performance and availability. The service makes it effortless for the developers as well as operators to improve application availability by automatically detecting operational issues early and recommending actions to take that can address the problem. It automatically ingests operational data from your AWS applications and provides a single dashboard to visualize issues in your operational data.
Amazon Monitorn is a managed service that is an end-to-end system that uses machine learning (ML) to detect abnormal behavior in industrial machinery. The system enables a user to implement predictive maintenance and reduce unplanned downtime.
Monitron includes sensors to capture vibration and temperature data from equipment. The Monitron service is a gateway device to securely transfer data to AWS that analyses the data for abnormal patterns using machine learning, and comes with a companion mobile app to set up the devices and receive reports on operating behavior and alerts to potential failures in your machinery.
Amazon Lookout for Equipment
Amazon Lookout for Equipment is AWS-managed ML a service that provides customers with existing sensors on their industrial equipment, a way to send their sensor data to AWS to build machine learning models for them and return predictions to detect abnormal equipment behavior. This enables predictive maintenance that allows them to take action before machine failures occur and avoid unplanned downtime.
With this automated machine learning tool, customers can bring in historical time series data and past maintenance events data generated from industrial equipment that can have up to 300 data tags from components such as sensors and actuators per model.
Amazon Lookout for Vision
Along with Lookout for Equipment, AWS also announced Lookout for Vision, a machine learning (ML) service that helps customers in industrial environments to detect visual defects on production units and equipment in an easy and cost-effective way. It uses deep learning models to replace hard-coded rules and handles the differences in camera angle, lighting, and other challenges that arise from the operational environment. With Lookout for Vision, you can reduce the need for carefully controlled environments.
During the event, Amazon previewed AWS Panorama, which is a machine learning appliance and software development kit (SDK) that allows organizations to bring computer vision to on-premises cameras to make predictions locally with high accuracy and low latency. One can now develop a CV model using Amazon SageMaker and then deploy it to a Panorama Appliance that can then run the model on video feeds from multiple networks and IP cameras.
The AWS Panorama Appliance is a hardware device that allows you to add CV to your internet protocol (IP) cameras that weren’t built to accommodate computer vision. Also, AWS Panorama Device SDK is a software kit that enables third-party manufacturers to build new cameras that run more meaningful CV models at the edge for tasks like object detection or activity recognition.