AWS re:Invent 2019 Geospatial Talks
Talks, sessions and workshops that may be of interest to those working with geospatial data. PRs accepted!
Video | Slides
Across the commercial and public sectors, companies are working with large geospatial datasets. We take a look at how to use various services including Amazon Simple Storage Service (Amazon S3), Amazon Simple Storage Service Glacier, Amazon Athena, AWS Step Functions, and AWS Batch to store, process, and get insights into your large geospatial datasets. Please bring your laptop.
Video | Slides
SpaceNet, a nonprofit LLC focusing on solving geospatial problems such as mapping road network routes after a natural disaster, has open-sourced more than 6,500 square kilometers of high-resolution satellite imagery with approximately 800,000 building footprint labels and 8,000 square kilometers of road network labels. Join us for a discussion on how to use this data to train machine learning algorithms that help disseminate timely information in the aftermath of natural disasters.
Video | Slides
In this session, learn how to design a data lake and how to give permission to different groups and applications to access and analyze datasets. Hear from subject-matter experts about a variety of AWS technology for populating your data lake, monitoring new ingestion, and processing data for meaningful analysis. We also examine considerations for structured data, such as relevant database engines with geospatial support, as well as considerations for unstructured data in the form of object storage. Finally, learn how to protect and secure data based on your organization’s needs.
Video | Slides
Launched in fall 2017, Bird is a micromobility company that enables access to shared e-scooters and lightweight electric vehicles in 100+ locations worldwide. Join us to hear how building a modern stack on top of Amazon EKS has enabled Bird to quickly ramp up its development in order to provide business value in a stable and secure manner. Further, learn how Bird’s backend utilizes native AWS services like Amazon S3 and Amazon SQS, open-source streaming systems like Kafka and Flink, and a modern microservices architecture to turn terabytes of geospatial data into the mobility revolution of the future.
Video | Slides
In this workshop, get hands-on experience managing large amounts of data in Amazon S3. Learn best practices for configuring and managing object policies, including storage buckets, security, regulatory compliance, data replication, and more. We review optimizing storage costs, data lifecycle management, and retrieval times through the various restore tiers and SLAs, and we share strategies for moving and organizing data in formats that take advantage of Amazon S3 features. We also discuss how efficient Amazon S3 clients use the Amazon S3 API. In addition, learn about listing objects, Amazon S3 inventory, Amazon S3 batch operations, and Amazon S3 Select.
Video | Slides
AWS hosts a variety of public datasets that anyone can access for free. Previously, large datasets such as satellite imagery or genomic data have required hours or days to locate, download, customize, and analyze. By making data available publicly on AWS, anyone can analyze any volume of data without needing to download or store it themselves. In this session, the AWS Open Data team shares tips and tricks, patterns and anti-patterns, and tools to help you most effectively stage your data for analysis in the cloud.
Video | Slides
Code Green invites you to enter a hackathon involving Amazon Sustainability Data Initiative (ASDI) datasets. Individual - or groups of two to five - hackers work on projects during re:Invent, with demos presented Thursday. Any ASDI dataset can be used. In particular, check out weather and climate. APIs, visualization tools, combinations of datasets into something new - all fair game. Entries should be open-source GitHub repos. Want to participate without competing? An ASDI workshop runs concurrently during the event Thursday. Code Green is purchasing carbon offsets for travel to re:Invent by all registered hackathon participants; hackathon winners receive free passes to re:Invent 2020. The real winner? Earth, of course.
Video | Slides
The BlueDot Water Observatory is an Earth-observation-based solution that provides reliable and timely information about surface water levels across the globe. Cost-effective yet reliable solutions for monitoring water resources are needed, as ground-based monitoring networks are often too costly and, in some cases, also unreliable. Sinergise shows how using global satellite imagery available on Amazon Simple Storage Service (Amazon S3) through the AWS Public Dataset Program, combined with an efficient use of services including AWS Lambda, Amazon DynamoDB, and Amazon CloudWatch, you can carry out a global-scale project cheaper than previously possible.
Video | Slides
Emergency personnel have to carefully coordinate their response to natural disasters across many teams. This coordination requires data, especially actionable mapping data. However, getting access to data at the tactical edge is challenging. In this session, we demonstrate an architecture and pipeline for managing data for field scenarios with the ruggedized AWS Snowball Edge. The serverless, cloud-based pipeline combines public and private data sources with open-source software that can be preloaded on Snowball Edge. See how it works firsthand, and ask questions to learn how you could put such edge computing to work in your field scenarios—even with drones.
Video | Slides
Many government agencies and commercial organizations come together to respond to natural disaster emergencies in the field. Such responders need computing resources at the tactical edge for communications, data gathering, and reconnaissance, often in dangerous or unstable environments. In this session, we discuss a solution architecture combining AWS Snowball Edge, IoT sensors, and C4ISR software that was deployed as the AWS Disaster Response Action team tested the tracking of search parties and recovery vehicles. We also demonstrate how machine learning at the edge can augment traditional sensor data to better support first responders and decision makers during disaster scenarios.
Video | Slides
In this session, build a processing chain in AWS using Amazon Elastic Compute Cloud (Amazon EC2), Amazon Virtual Private Cloud (Amazon VPC), and Amazon Simple Storage Service (Amazon S3) to take weather satellite imagery data from the AWS Ground Station service and process it to finished images in Amazon S3. Learn about the different streaming data formats that AWS Ground Station provides and the steps involved in processing that data through different layers and components such as reliable delivery, demodulation and decoding, data recovery, and developing an output product. Please bring your laptop.
Video | Slides
Amazon SageMaker Ground Truth helps you build and manage highly accurate training datasets quickly. Ground Truth offers easy access to public and private human labelers and provides them with prebuilt workflows and interfaces for common labeling tasks. During this builders session, we show you how to start and use a workflow. Please bring your laptop.
Video | Slides
Join us for a technical deep dive that provides you training to build a mission profile using AWS CloudFormation. In this session, you learn the satellite onboarding process for AWS Ground Station, build a mission profile that properly configures the antenna system before the pass, and directs your data flows between the antenna system and your VPC. You even get to watch a satellite contact! Please bring your laptop.
Video | Slides
Successful machine learning models are built on high-quality training datasets. Typically, the task of data labeling is distributed across a large number of humans, adding significant overhead and cost. This session explains how Amazon SageMaker Ground Truth reduces cost and complexity using techniques designed to improve labeling accuracy and reduce human effort. We walk through best practices for building highly accurate training datasets and discuss how you can use Amazon SageMaker Ground Truth to implement them.
Video | Slides
NASA Jet Propulsion Laboratory’s (JPL) IT Chief Technology Officer, Tom Soderstrom, presents a demonstration of automated scheduling with AWS Ground Station and a NASA JPL satellite. The satellite, ASTERIA, used for this demonstration was designed in collaboration between the Massachusetts Institute of Technology and NASA JPL. AWS Ground Station connects antenna systems to cloud technologies so that researchers and scientists can automate their projects in space.
Video | Slides
The first hour after a natural disaster is often referred to as the “golden hour," when responders have the highest chance of saving the lives of those effected by the disaster—but it's also when they have least amount of information available to them. In this chalk talk, we walk through different ways machine learning can help accelerate a responder’s understanding of the areas impacted and their ability to begin formulate potential recovery strategies. We also discuss how we leveraged Amazon SageMaker, In-Q-Tel’s SpaceNet dataset, Amazon Comprehend, AWS Lambda, Amazon API Gateway, and much more to build the prototype solution.
Video | Slides
With AWS Lambda, you can write code to process images and files without provisioning servers. But at scale, how do you coordinate multiple parallel processing steps, catch errors and retry failures, and keep your code modular and maintainable? AWS Step Functions comes to the rescue. In this workshop, you design and implement a distributed state machine to orchestrate a multi-step image recognition and processing workflow using Amazon Rekognition and AWS Step Functions.
Video | Slides
Amazon SageMaker Ground Truth makes it easy to quickly label high-quality, accurate training datasets. In this workshop, we set up labeling jobs for text and images to help you understand how to make the most of Amazon SageMaker Ground Truth. You learn how to explore and prepare the dataset and label it with object bounding boxes. Then, we use Amazon SageMaker to train a Single Shot MultiBox Detector (SSD) object-detection model based on the labeled dataset, use hyperparameter optimization to find the best model for deployment, and deploy the model to an endpoint for use in an application.
Video | Slides
Amazon Athena is an interactive query service that makes it easy to analyze data in Amazon Simple Storage Service (Amazon S3) using standard SQL. Athena is serverless, and customers pay only for the queries they run. In this workshop, we dive deep into various use cases for Athena, including building applications that run on schedules, using CTAS and INSERT INTO as effective self-service ETL tools, and querying storage formats. We also look at authorization, authentication, and managing costs. If you are a data engineer looking to onboard your organization to Athena, this workshop equips you with the tools to be successful.
Video | Slides
Learn how Maxar and Descartes Labs run complex, global-scale models on Amazon EC2 instances powered by Intel Xeon Scalable processors. Maxar discusses its experience setting up an operational HPC cluster to run global numerical weather prediction models, obtaining performance that eclipses the speed of the NOAA bare-metal supercomputer. Descartes Labs shows how its platform enables hyper-scale object detection on satellite imagery accelerated by Intel AVX-512 instructions. It also shares its experience deploying tightly coupled HPC applications that use spot blocks at many-thousand processor scale, using HPC clusters built on AWS instances and powered by Intel Xeon Scalable processors. This presentation is brought to you by Intel, an APN Partner.
Video | Slides
ALERTWildfire is a camera-based network infrastructure that captures satellite imagery of wildfires. In this chalk talk, we discuss deep-learning techniques that use this satellite imagery along with meteorological data to track wildfires and predict air quality in real time.