Find out more about why attending re:MARS makes sense for you.Why attend
Check out all the sessions available at this year's Amazon re:MARS. Register today to start favoriting sessions and planning your time at re:MARS.Learn more
For a limited time only, register with discount code DG150QqVRkT2a to get a $150 discount on your full conference pass to Amazon re:MARS.Register now
Join us in Las Vegas to get inspired by the latest in MARS technologies, learn career-changing new skills, and become part of a community that will shape the future of the world.Learn more
More industry-recognized speakers have been added to the re:MARS 2022 line-up, including Alicia Boler Davis, SVP of Customer Fulfillment at Amazon.Learn more
We have updated our mask and vaccine requirements for re:MARS. Please review our Health measures page for the latest information.Learn more
Join this session to learn how AstraZeneca is driving insights at scale and putting the power of artificial intelligence (AI) in the hands of employees by activating self-service capabilities on AWS, helping them deliver life-changing medicines. Discover how AstraZeneca has implemented a new integrated machine learning (ML) platform across the enterprise and is now expanding to the smart factory digital transformation program to reduce the time it takes to discover and manufacture a potential drug candidate. Learn how AstraZeneca hosts ML models for prediction and inference, generative AI, and accelerated drug discovery in clinical trials, all at enterprise scale.
In 2019, Amazon and the National Science Foundation (NSF) announced a $20 million collaboration to fund academic research on fairness in AI over a three-year period. In this session, hear from a panel of select grant recipients of the NSF Fairness in Artificial Intelligence (FAI) program and Amazon researchers on their perspective on FATE topics applied to machine learning, automation, robotics, and space (MARS) themes.
When you’re deploying an ML model update, A/B testing of old and new models using production data improves accuracy and delivers better outcomes. In this session, learn how you can quickly and easily instrument your Amazon SageMaker applications using Amazon CloudWatch Evidently for A/B testing and gather meaningful statistical insights without advanced training to interpret the data. Learn how AWS Activate improved its ML models for personalization to offer smarter predictive choices for its customers.
Behind every drug in the market, there are thousands of compounds that get discarded along the drug development pipeline. Knowing what is going to work before it gets produced is the ultimate goal of drug development. In this session, hear about some of the latest work the Amazon Machine Learning Solutions Lab, in collaboration with customers, has done to get closer to this important goal. Walk through key use cases, solutions, and state-of-the-art approaches that are emerging in this area.
To make decisions about a single patient, today’s clinicians must reconcile growing amounts of diverse health data from multiple sources. Pranav Rajpurkar’s research cuts across computer vision, natural-language processing, structured health data, and medical imaging to accelerate and improve the clinical decision-making process. In one study, this research helped reduce the time required to interpret chest X-rays from 240 to 1.5 minutes. By better predicting the healthcare needs of patients in vulnerable populations, the work also resulted in savings of $200 a year per patient. In this session, learn how artificial intelligence can help empower the next generation of clinicians.
Human transcriptions are slow and costly and therefore cannot provide full coverage to all Amazon Alexa applications nor keep pace with evolving usage. In this session, hear about a real-time continual, lifelong learning system that trains ML models using production data at scale, without persisting data or retraining from scratch. Its self-learning loop leverages intrinsic model-inferred signals such as confidence and external signals such as customer friction to learn new usage patterns, adapt to concept drift, label data without human supervision, and correct errors. It combines this self-supervised online data with synthetic or historic human-labeled data to continuously improve models.
In this session, learn about diversity, equity, and inclusion, and how machine learning, computer vision, and artificial intelligence can change the world for differently abled people with an American Sign Language interpreter. Discover how existing ML models like image classification and neural networks can help to build such a model. The session starts with the basic concepts, covers data transformation, and discusses model training and deployment for our use case.
Ambient Explorer for Shopping's mission is to make Alexa the world’s most knowledgeable shopping assistant, inspiring and informing customers’ shopping journeys using content from the broader web. Alexa customers shopping for gifts will be able to find inspiration via content from web publishers conversationally without having to go between Amazon and other sites. The conversational experience is powered using Conversation Mode and Alexa Conversations. In this session, learn how Alexa Conversations uses AI to bridge the gap between experiences developers can build manually and the vast range of possible conversations using dialog expansion methods.
The future of human advancement depends on the deployment and scaling of technology to every person on the planet. In this session, learn how Intel and its partners bring transformational machine learning, automation, robotics, and space (MARS) platforms to users and educate next-generation programmers, developers, and entrepreneurs, enabling them to harness the power of silicon. This presentation is brought to you by Intel, an AWS Partner.
This session provides an overview of how AWS deployed PIPE AI products. Learn how Predict generates asset failure predictions and Review ingests CCTV footage to automate the analysis of sewer, stormwater, and rooftop defects. We provide a breakdown of MLOps on AWS, which enabled delivery of an end-to-end solution, pipeai.com.au, for utilities to better predict and identify failure. Finally, you learn about some of the challenges in embarking on prediction and computer vision projects, including data capture, curation, and engineering, as well as some of the gaps between historical human-driven regulatory defect detection requirements and machine-driven outcomes. This presentation is brought to you by Blackbook.ai, an AWS Partner.
Conversational AI is the next frontier in elevating customer experiences, irrespective of industry or use case. State-of-the-art innovations are delivering multi-modal solutions that are adaptive and predictive and that use data-driven decisioning. In this session, meet Midas, a virtual assistant for adverse event detection in a healthcare setting, and TrueServe, a low-code/no-code advanced AI next-generation contact center platform with hyperpersonalization and prebuilt libraries of industry-agnostic use cases. Our specialists take you behind the scenes on how you can integrate CRM, SMS, phone, voice, web chat, and more to deliver high-quality customer engagements using the latest machine learning and AI toolkits. This presentation is sponsored by Deloitte, an AWS Partner.
Climate researchers have increasingly adopted AI/ML to tackle complex challenges, including prediction, process parameterization, and knowledge discovery. In this session, explore how scientists are using full-scale climate models, explainable AI, and physics-informed AI models to facilitate a deeper understanding of climate processes, climate impacts, and the potential impacts of climate intervention techniques designed to mitigate the worst effects of global warming, such as increasing the reflection of sunlight from particles in the atmosphere. Join the session to understand how AWS HPC clusters and training networks on GPU nodes are pushing the frontiers of climate science and climate-risk research.
Amazon Alexa’s response model was created for headless devices with one succinct answer to a user request. In this session, learn about blended results, a combination of results from first-, second-, and third-party skills, which expands the traffic footprint for all developers on multimodal devices. Requests like “Alexa where can I get pizza?”, are answered with three options for selection/disambiguation (shop pizza from WholeFoods, order Dominos, and pizzeria from local search). Relevant follow-up actions are offered after answering “Alexa, how can I lose 100 calories?”, by placement of fitness cards like “7-Minute Workout” and “Easy Yoga” to predict next action.
In this session, a panel of venture capital firms, including AI Fund, Sequoia Capital, and Madrona Venture Group, as well as machine learning (ML) founders from Tecton, WhyLabs, OctoML, and ValidMind discuss the latest trends, open frameworks, and growth in this emerging segment. This session is moderated by Vin Sharma, AWS GM of ML Edge and Engines.
In this session, learn about the science of extracting information from scanned documents and how it is applied in the new service feature Amazon Textract Queries, which is helping banking and insurance customers provide faster service by accelerating loan processing, claims reimbursement, and more. Hear directly from a customer who is already benefitting from this automation.
Exploratory data analysis (EDA) is a critical step in the machine learning (ML) journey. Data scientists must explore their datasets to understand them, clean them, and perform feature engineering. In this session, consider how EDA is rapidly changing with the introduction and evolution of tools that prepare data for ML without requiring users to write code. Dive deep into two features of Amazon SageMaker Data Wrangler: automated insights into data quality and feature engineering and the ability to generate hundreds of features on time-series data with a single click.
A key challenge working with real-world healthcare and life sciences (HCLS) patient data is the siloing of data across multiple hospital systems and research facilities. Regulations prohibit open data sharing, and the complexity and cost of centralized data repositories deter their use. HCLS partners and customers seek privacy-preserving mechanisms for managing and analyzing distributed and sensitive data. In this session, explore a proposed federated learning framework built on AWS, which facilitates training a global machine learning model on distributed healthcare records. Learn about the effectiveness of the framework on eICU, a critical-care database with over 200 contributing hospitals.
In this session, learn how Amazon is developing common-sense reasoning over Amazon Alexa's knowledge of the world, device states, and customer preferences. For example, by reasoning about connected-appliance state, Alexa can automatically recover from errors such as recognizing “turn on lights” (instead of “turn off lights”) when the lights are on. Alexa can answer questions like “Can the Seahawks make the playoffs?” based on results, scheduled games, and knowledge of the qualification process. Alexa can proactively take time-dependent actions, such as adding missing recipe items to a shopping list or playing music based on users’ listening habits, situation (e.g., walking or running), and time of day.
Building and training of ML workloads can be an energy-intensive process, and high accuracy and large computational resources necessitate substantial energy consumption. In this session, explore guidance from the sustainability pillar of the AWS Well-Architected Framework to learn how to reduce the carbon footprint of your AI/ML workloads. The session covers best practices for efficiently retraining multiple models using minimal computational resources and computationally efficient built-in algorithms. Also learn about the AWS tools available for monitoring your model during training and deployment.
The frequency and severity of natural disasters are increasing. In response, governments, businesses, nonprofits, and international organizations are placing more emphasis on disaster preparedness and response. Many organizations are accelerating their efforts to make their data publicly available for others to use. Repositories such as the Registry of Open Data on AWS and Humanitarian Data Exchange contain troves of data available for use by developers, data scientists, and machine learning practitioners. In this session, see how a community of developers came together though the AWS Disaster Response hackathon to build models to support natural disaster preparedness and response.
In this session, learn how AWS solves robot trajectory planning problems at industry-relevant scale. The AWS end-to-end solution integrates random-key algorithms with path relinking for solution refinement. Through a distinct separation of problem-independent and problem-dependent modules, learn how to achieve an efficient problem representation with a native encoding of constraints. See numerical results for industry-scale datasets and how the approach consistently outperforms the greedy baseline. Hear why, to assess the capabilities of today’s quantum hardware, we complement our classical approach with results obtained on quantum annealing hardware via QBSolv, providing a path for a quantum-ready hybrid solution.
Are you interested in how machine learning (ML) and robotics can revolutionize industrial processes? Are you curious about using images to build custom ML models? Attend this session to learn how Baxter International, Inc. uses Amazon Lookout for Vision to build custom computer vision models to automate their quality inspection to detect defects in their manufacturing process while reducing waste and improving operating margins.
As financial institutions adopt AI/ML, the regulatory requirements are mandating explainability of the machine learning models. This requirement has led financial institutions to rely heavily on the interpretable models built using linear regression, logistic regression, and decision trees. In this session, learn about the development, operational, and process improvements that can be incorporated by organizations to improve the explainability of models while adhering to the regulatory requirements.
One of the most active areas in AI research is the use of synthetic data, where simulations are used in place of real-world sensor collection to train AI algorithms. This may look like a straightforward process of running simulations in bulk. In reality, successful creation of synthetic data requires additional technologies, broad arrays of content, novel data pipelines, AI-centric quality metrics, and workflows that are relevant to a new (and very different) class of users. In this session, hear about some of the leading work being done in synthetic data with an eye toward how these gaps can be closed.
Learn how GovChat, the official citizen engagement platform for the South African government, is transforming citizen engagement with the government using machine learning. To date, over 8.7 million South African citizens are using GovChat to communicate with the government regarding critical and essential services during COVID-19 pandemic. The GovChat technology platform uses a chatbot to allow real-time feedback and reporting through its mobile device and has processed over 582 million messages over the past 2 years. Learn how GovChat is leveraging AWS services such as the advanced natural language capabilities of Amazon Lex to build this chatbot and improve citizen experiences.
How can users scale large machine learning (ML) workloads on Amazon EC2 and PyTorch? In this session, review architectures and software stacks to scale ML model training with PyTorch. See the usage of fully sharded data parallel (FSDP) to train 175 billion- and 1 trillion-parameter models using parallel cluster/SLURM, and learn how Amazon EKS and TorchElastic can be used to train large models in a fault-tolerant manner. Learn why AWS services like Amazon EC2, Amazon EKS, and EFA are key ingredients to a scalable ML architecture, and learn from the success stories of customers who have scaled ML training across thousands of GPUs on AWS and PyTorch.
In American football, players can sustain head impacts during play, but the accumulated effect of these impacts to players’ health is not fully understood because manually identifying impacts is prohibitively time-consuming and inconsistent across individuals. In this session, learn how the NFL is working with AWS to develop a machine learning system that combines player tracking data with video footage to detect helmet impacts and associate them with the correct player. The system is capable of analyzing full NFL seasons over 10 percent more accurately and over 150 times faster than humans. The complete impact data drives insights to help the NFL improve player safety.
Speech recognition technology is dramatically changing by moving to a single neural architecture replacing the conventional stack of disjointly-trained neural and non-neural subsystems. In this session, learn how single neural architecture improves accuracy, while achieving superior memory and compute compression, making streaming low-latency speech recognition possible at the edge. The single neural architecture overhauls the boundary between speech recognition and language understanding, allowing for a jointly-trained unified stack which improves accuracy for both tasks. Progress in unsupervised or semi-supervised learning, combined with the single neural architectures provides leeway for future accuracy improvements, while reducing the costly reliance on human supervision.
Join this session to learn about the mathematics that provide the foundation for machine learning, unification of various machine learning methodologies, and moving toward the creation of models which are simultaneously precise and explicable.
As the adoption of AI assistants increases, it is pertinent that the AI technologies address and help alleviate diversity-related gaps. In the past two years, Amazon has launched and activated several features for Amazon Alexa that work to address diversity gaps. In this session, hear details of how the Alexa team identified these gaps and then invented new ML techniques to address them. Dive deep into the concept of speech disentanglement and how the team uses this technology to influence different aspects of speech—tone, phrasing, intonation, expressiveness, and accent—to create unique responses that make Alexa’s magic work optimally for everyone.
In this chalk talk, learn how we can utilize out-of-the-box ML capabilities to deliver a solution for a use case. Explore how you can take ML capabilities out of their trained domain and apply them to new use cases. Also see a demonstration of a map analysis tool.
Proteins are the main engines that drive the biological ecosystem. Protein structure sets the foundation for its function, and accurately predicting protein structures can improve and open up many downstream applications, such as drug development. In this chalk talk, learn what AlphaFold, the state-of-the-art protein structure prediction model, can do and possibilities beyond AlphaFold.
Not having a sufficient amount of training data can be an issue when developing computer vision models, especially when the data you’re collecting is rare and must be in multiple lightings and qualities. Typically, the solution has been to collect more data or use techniques like generative artificial intelligence to make fake data. Recently, gaming engines like Amazon Nimble Studio and O3DE have demonstrated the ability to generate realistic scenes and objects that can be used in machine learning training. In this chalk talk, learn how to use a gaming engine to generate training data for computer vision applications.
Mega AI aims to deliver cutting-edge, next-generation AI capabilities unique to the Department of Energy national lab complex. Mega AI addresses existing research gaps in narrow AI, specifically in scaling AI to hundreds of billions of parameters. In this chalk talk, learn about Mega AI, how it can deliver solutions for government researchers in multimodal representation learning, multitasking general-purpose inferences, the need for increased generalizability, the rapid development and deployment of AI technologies, and the usability and assurance of AI models for science and security applications.
The use of specialized processors dates back to the 1970s and 1980s when CPUs were paired with a math coprocessor. Today, a machine learning (ML) practitioner can speed up ML with an AI accelerator—a dedicated processor designed to accelerate ML computations. Users can choose from GPUs, AWS Trainium, or Habana Gaudi to accelerate training and can utilize GPUs or AWS Inferentia to accelerate ML inferences. In this chalk talk, discuss what an AI accelerator is, how it works, when to use it, and which to select based on performance and budget priorities.
In most enterprises, the machine learning lifecycle becomes stagnant after deployment. Retraining is either completely nonexistent or performed on a fixed schedule. This chalk talk dives into model monitoring mechanisms and how to automate retraining to gain efficiency and performance for Amazon SageMaker-hosted models.
Companies want effective tools that are simple to operate, and voice is one of the best tools for when an operator’s hands are occupied. While conversational AI typically requires a connection to the cloud, NVIDIA's Riva SDK activates true conversational AI at the edge. When paired with AWS IoT Greengrass V2 and low-power wide-area network, the ability to use voice to control irrigation systems and query precision telemetry data across large areas is achievable without a connection to the cloud. In this chalk talk, consider smart irrigation with the connected farm and explore how to take advantage of NVIDIA Riva.
Data silos can create barriers to efficiency and data quality and, as a result, often hinder an organization’s ability to begin their data and ML journeys in the cloud. Manual data ingestion, data preparation, and application integration are time-consuming and expensive. In this chalk talk, learn how organizations can efficiently break down data silos with low-code/no-code integration services such as Amazon AppFlow and speed up innovation with the help of AWS ML services. Walk away with prescriptive guidance on how to select the best data integration tool for a range of data and ML scenarios.
AI/ML can help improve several processes in the oil and gas industry, from predicting safety issues to optimizing exploration, drilling, and production. Today, much of the analysis in exploration is done manually, which is tedious and error-prone. In this chalk talk, see how Amazon AI services can automate tasks to extract deeper insights to improve decision-making and reduce interpretation time from months to days. Engage in a discussion about additional applications for ML in oil and gas, including improving safety outcomes, improving asset management and maintenance, and optimizing well placement.
With the growth in IoT, self-driving cars, drone technology, and industrial ML, it’s essential to bring ML to the edge. However, customers who struggle with ML may reject bringing solutions to the edge due to the additional complexity of doing so. In this chalk talk, learn how to simplify the journey and bust the myth of ML difficulty with example use cases. Also get hands-on with a drone and embedded devices that combine AWS IoT and Amazon SageMaker services.
Federated learning (FL) is a distributed machine learning (ML) approach that allows you to train ML models on distributed datasets. The goal of FL is to train better ML models with more data, while preserving the privacy of distributed datasets. FL can be applied to various vertical industries, including finance, telecommunications, healthcare, and IoT. In this chalk talk, learn about a cloud-native FL platform that activates MLOps for FL using AWS services and supports customized FL algorithms. The objective of the platform is to simplify the initial FL configuration, model training, and model deployment.
Many organizations are interested in improving their supply chain solutions using AI/ML. ML techniques can be used to solve challenges related to uncertainties in markets, unexpected increase or decrease in demand, and traceability and transparency in operations. The COVID-19 pandemic has intensified supply chain issues due to reduced workforce and increased remote work, so developing sustainable solutions for improving and automating supply chain processes has never been more important. In this chalk talk, explore use cases and successful customer stories demonstrating how ML can be used to solve supply chain challenges that businesses are facing today.
Many organizations are facing the challenge of efficiently extracting information hidden within complex network structures—for example, identifying fraudulent claims in healthcare through detecting abnormal relationships between patients and providers, detecting abnormal financial transactions between different entities as an anti-money laundering tool, or segmenting an audience for targeted marketing campaigns. In this chalk talk, explore how hybrid quantum annealing on Amazon Braket can help augment traditional AI/ML models to solve complex community detection problems.
This session discusses inherit biases in AI/ML. Learn how diversity, equity, and inclusion (DEI) intersect with AI/ML to help build the future of technology.
Developers, start your engines! AWS DeepRacer is the fastest way to get rolling with machine learning (ML), literally. This workshop provides developers of all skill levels an opportunity to get hands-on with AWS DeepRacer to learn the basics of reinforcement learning, an advanced ML technique. During this workshop, dive into the AWS DeepRacer console to build a reinforcement learning model for an autonomous driving application that is ready to race in under 90 minutes. Take your model from the classroom to the track and compete in the AWS DeepRacer League for prizes and glory.
Today’s business users rely on business intelligence teams to gain insights into their data, which can bottleneck time-sensitive decision-making. Amazon SageMaker Canvas and Amazon QuickSight give you the power of ML without any code or ML experience to gain data insight when you need it most. In this workshop, learn how to build dashboards from your data quickly without reliance on specialized teams. Learn how to build and use machine learning models in minutes to hours for rapid prototyping and validation of ideas.
In order to produce better models, datasets are exploding in size which is leading to longer training times. Amazon SageMaker Distributed Data Parallel library helps you scale training to multiple instances in a way that uses SageMaker's resources more efficiently. In this hands-on workshop, learn how you can quickly and easily modify a model and minimize its training time using SageMaker Distributed Data Parallel library.
In many real-world applications, anomalous data is unlabeled, which makes it challenging to detect without labeled examples. This hands-on workshop shows you how to use Amazon Lookout for Equipment and other AWS services to build an automated pipeline from anomaly detection to predicting failures with multivariate time series analysis for industrial and manufacturing use cases (sensors, equipment, and process data). Learn how the same principles can apply to spacecraft telemetry and subsystems using autoencoders—a powerful encoding technique to separate anomalous and nominal data.
Millions of applications and hundreds of millions of forms (think W-2s) are processed by the U.S. government each year, mostly using manual data entry. Amazon Textract and Amazon Comprehend automate document processing by combining optical character recognition (OCR) and natural language processing (NLP) for faster results with higher accuracy. In this workshop, learn how to architect an ML solution to extract text and data from these types of documents at scale for automatic processing. Discover how to build a serverless, highly available, scalable architecture that can handle spiky workloads.
The amount of data in the world is growing at an exponential rate and for enterprises across almost every vertical (e.g., IoT, manufacturing, finance, healthcare, etc.) that includes time series and JSON-formatted data. In this workshop, explore various techniques to incorporate the power of these data types into your ML pipeline using Amazon SageMaker Data Wrangler. Learn how to incorporate new types of data such as time series and JSON data into machine learning models, which can dramatically improve model performance and bring game-changing insights.
Amazon Personalize is a fully managed ML service that goes beyond rigid, static, rule-based recommendation systems and trains, tunes, and deploys custom ML models to deliver highly customized recommendations to customers across industries such as retail, media, and entertainment. In this workshop, explore multiple approaches for creating personalization engines on AWS using Amazon Personalize and customer personalization techniques using Amazon SageMaker.
As LiDAR sensors become more accessible and cost-effective, customers are increasingly using point cloud data in new spaces and for ML tasks like 3D object detection and tracking, 3D segmentation, 3D object synthesis and reconstruction, and use of 3D data to validate 2D depth estimation. This workshop features Amazon SageMaker Ground Truth and explains how to ingest raw 3D point cloud data, adjust and label it, train a 3D object detection model, and deploy this model to an endpoint. Learn how to train your model on an autonomous vehicle dataset using techniques that you can apply to most 3D point cloud data.
GPT-J-6B is an autoregressive language model that performs well on a wide array of natural language processing (NLP) tasks. For this reason, GPT-J-6B has recently generated a lot of interest from researchers, data scientists, and software developers, yet it remains very challenging to fine-tune the model because it is hard to fit it into a single GPU device. In this workshop, learn how to fine-tune a GPT-J-6B model using Amazon SageMaker’s model parallel library and deploy the model on SageMaker for inference, a solution adopted by multiple enterprise companies for their NLP needs.
Billions of IoT devices will connect with the cloud over the next five years. With IoT continuing to converge with AI, robotics, and 5G, enterprises are tackling big problems in infrastructure, production, and the environment using cloud technologies. But how do you get started and separate dream from reality? Through demos and customer stories, Michael MacKenzie, GM of Industrial IoT & Edge at AWS, introduces audiences to the new IoT and robotics technologies that are transforming building management, operations, and manufacturing, and discusses how customers like Volkswagen, Carrier, and Yara are increasing plant efficiency, automating processes, and acting upon real-time insights at scale.
Digital twins represent an exciting new technology that businesses are adopting to make better operational and strategic decisions in industries such as automotive, aerospace, manufacturing, healthcare, life sciences, energy, power and utilities, and other industrial operations. Digital twins require the convergence of at-scale computing, new machine learning methods, spatial computing (3D/AR/VR), and IoT connectivity. In this session, learn how AWS can help customers understand their use cases and the technologies needed to achieve their business goals to scale their digital twin journey. Also discover how Siemens Energy is developing digital twin solutions on AWS for their operations.
Digital twins represent an exciting new technology that businesses are adopting to make better operational and strategic decisions in industries such as automotive, aerospace, manufacturing, healthcare, life sciences, energy, power & utilities, and other industrial operations. In this session, see how AWS can help customers understand their use cases and the technologies needed to achieve their business goals to scale their digital twin journey. Then, hear how MHP is developing a digital twin solution to predict the performance of electric vehicles with a focus on vehicle-specific battery degradation dependent on driving patterns.
Amazon expands the frontier for speed of delivery and selection access with an understated achievement: a large-scale, hands-off-the-wheel automated system that uses advanced mathematical optimization and machine learning. In this session, learn how Amazon innovated and deployed one of the largest algorithmic decision-making machines in the world, how it has helped Amazon more than triple its scale in five years, and how Amazon went back to the drawing board to develop new science to manage and optimize its increasingly complex multi-echelon network.
A small startup in Bangladesh, SOLShare has created smart “microgrids” that are tackling climate change one village at a time, and bringing affordable renewable solar electricity to communities in Bangladesh and beyond. In this session, dive into the AWS IoT technology that supports the world’s first peer-to-peer solar sharing grids, and share stories from the rural communities that are being empowered. The peer-to-peer aspect is fascinating: people can bank extra solar power and sell it to neighbors through mobile money wallets.
Fully autonomous driving is the future and on its way to revolutionizing mobility. Achieving acceptable levels of safety and reliability requires significant verification and validation using enormous data volume generated through continuous road tests for diversified scenarios. In this session, learn how Capgemini’s Driving Automation Systems Validation (DASV) helps automotive manufacturers build autonomous vehicle development strategies for advanced driver-assistance systems (ADAS) with end-to-end data management and AI capabilities. DASV provides reduced time to data and enhanced testability using AWS services, including AWS IoT Core, AWS Outposts, AWS Snowball, AWS Wavelength, Amazon S3, Amazon RDS, and Amazon EKS.
The central innovation team at Novo Nordisk has developed an industry-first automated quality assurance system using AWS services to count drug cartridges in real time on the manufacturing line. In this session, learn how they developed a demonstration in which a robotic arm places a box full of drug cartridges on a platform, a camera rig takes images of the box, ML inference is performed at the edge (using an edge device), and the final results are displayed on an Amazon QuickSight dashboard.
A majority of trained models never make it to production and those that do take an average of 12 weeks to be deployed. To tackle these issues, organizations are establishing best practices to treat models like software much earlier in their lifecycle. This allows organizations to integrate their ML quickly with existing DevOps tooling/workflows and AWS deployment environments. In this chalk talk, learn strategies for how to create a seamless handoff between data science teams and IT operations, so your models can boost the intelligence of cloud-native applications running in AWS container services.
Replicating a common use case from the energy industry, this chalk talk demonstrates using a real water tank and connecting sensor data to AWS IoT services so that the data is available in the AWS Cloud. Through this demo, discover how to create a digital twin of real-world systems using AWS IoT TwinMaker to connect, combine, and present real-time IoT data within an immersive 3D visualization for easier analysis and understanding. Additionally, see how to generate alerts when a water leak has been automatically detected in the water tank system.
IT and security teams want to help their data science teams rapidly develop and iterate on models rather than have them worry about how to set up, govern, and maintain the configurations of their AWS resources. In this chalk talk, learn how to automate the deployment of a governed and secure data science environment for your team from a managed AWS Service Catalog deployment using AWS CDK. With this approach, your data science teams will be able to launch ML and analytical services, including Amazon SageMaker Studio and Amazon EMR Studio across a multi-account environment.
Enterprise users adopting machine learning (ML) on AWS often look for prescriptive guidance on implementing security best practices, establishing governance, securing their ML models, and meeting compliance standards. Building a repeatable solution provides users with standardization and governance over what gets provisioned in their AWS account. In this chalk talk, learn steps you can take to secure third-party ML model deployments. Explore cloud infrastructure-as-code templates that automate the setup of a hardened Amazon SageMaker environment. These templates include private networking, VPC endpoints, end-to-end encryption, logging and monitoring, and enhanced governance and access controls via AWS Service Catalog.
Extreme weather events and poorly managed forests are causing unprecedented wildfires globally. Every year, utility companies inspect thousands of miles of transmission lines in search of vegetation at risk of contacting lines and causing wildfires. Leveraging deep learning on satellite images and LiDAR data using AWS machine learning services can identify areas of risk. Utility companies can use the identified anomalies to monitor vegetation and proactively intervene to prevent wildfires and protect critical infrastructure. In this workshop, learn how to use Amazon SageMaker to process satellite images and LiDAR data and identify vegetation risks using deep learning.
Methane emissions can be harmful to the environment and dangerous to field workers. Join this workshop to learn how to build a simple gas detector using a Raspberry Pi and worker badge with an AWS IoT EduKit. Provision an IoT device on AWS, stream real-time data to the cloud, and then visualize the results in a dashboard. Telemetry data can be hard to conceptualize, and this workshop grounds the data collection in the actual process of connecting data with an edge device and demonstrating how that data can be used to identify methane leaks in real time and react to them.
This workshop introduces participants to IoT and ML technologies that can extract insight from industrial data and use it to improve industrial processes. Gain hands-on experience using AWS IoT Greengrass to deploy and run software on edge devices. Learn how to add an edge device to an industrial control network and publish sensor data to the cloud for near real-time monitoring and analysis of historical data. This historical data is then used to build ML models for predictive (what-if analysis) and prescriptive (set point optimization) purposes. Finally, deploy ML models to edge devices and use insights as part of latency-critical applications.
Join this chalk talk to discuss defect detection at the edge with Amazon Lookout for Vision. Explore how to get good results, learn from a variety of use cases, and dive deep into both the larger picture of anomaly detection and its smaller details.
Robots are increasingly becoming a vital component in a number of industries. Robots are able to complete more complex tasks at an impressive rate. This session discusses the anticipated future direction of and future fields of application for robotics. Learn how to identify if robotics is the right solution for you, and hear about some upcoming opportunities and challenges in autonomous robotics.
Agriculture is one of the least digitized industries. Every year, there are trillions of pieces of fruit picked by hand. Farmers continually struggle to find and manage people, which results in huge levels of waste and lost profit. Ripe Robotics was founded with the goal of revolutionizing an industry that has been left behind by technological innovation. Fruit has been picked largely the same way for much of human history. Join this session to learn how Ripe Robotics is using recent advances in artificial intelligence and robotics to make automation possible in this field.
Robots require an intricate orchestration of many complex parts, from perception and world modeling to planning and control. NVIDIA’s Isaac Sim, built on the Omniverse development platform, exposes a breadth of tools for synthetic data generation, world representation, RL training, planning, and other features needed to express sophisticated reactive robotic behavior for next-generation collaborative robots. In this session, learn how Isaac Cortex orchestrates these tools into a simple programming framework for design and deployment of complete collaborative systems. This presentation is brought to you by NVIDIA, an AWS Partner.
Robots are becoming an indispensable tool in space exploration. Traditionally, these systems were developed using custom, proprietary software written for each application and hardware platform. Meanwhile, the broader robotics community has standardized on ROS, an open-source platform for building robotics applications. Now the space industry is following suit by developing Space ROS, which will allow ROS to be qualified for space applications and reused across missions, saving development effort and speeding up robotic space exploration. In this session, hear about the past, present, and future of space robotic exploration and discuss the plans for Space ROS.
Turner Topping, Robotics Engineer at Ghost Robotics, shares how the company’s mission of building the most rugged and easiest to deploy legged robots is powered by 5G edge processing and AWS Snowcone.
Industry 5.0, broadly known as industrial automation, refers to people working alongside robots and smart machines. It’s about robots helping humans work safer, smarter, and faster by leveraging advanced technologies like the Internet of Things (IoT) and big data. This $300 billion market intersects with AI, ML, robotics, supply chain logistics, and warehousing. As of 2021, over $25 billion has been invested by venture capitalists into over 1,000 startups. The newly launched Amazon Industrial Innovation Fund seeks to help grow and accelerate startup activity. In this session, explore why startups could benefit from collaborating with the Amazon Industrial Innovation Fund.
Globally, sustainable agriculture techniques are critical for future food security. Sustainable agricultural development faces many obstacles such as cultivated land loss, injudicious use of fertilizers and pesticides, and environmental degradation. In this session, hear how Agrointelli is building automated robots to preserve and improve the soil functions needed to handle both climate changes and a growing population. Learn how Agrointelli built a fleet of failure-resistant robots that can be easily updated for changing environmental and soil conditions throughout the harvesting season.
Building management applications that work across a fleet of robots is challenging. As a result, many companies manage these robots independently, in silos, making it difficult to orchestrate tasks and monitor robot location on a single map. In this session, learn how AWS IoT RoboRunner makes it easier for developers to build applications that require robot interoperability such as task allocation and space management. Explore how to use AWS IoT RoboRunner to unify robot data and control systems into a single, central repository and then build applications to view and orchestrate robot tasks within a fleet.
Companies like Amazon rely on robots to automate their operations in warehouse facilities, fulfillment centers, manufacturing lines, and more. Simulation is a critical tool to accelerate robotic solution delivery, train robots more accurately, and optimize complex robotics operations before physical deployment. In this session, learn how Amazon Robotics uses simulation, synthetic data generation, and 3D digital twins in all phases of the robotic project development lifecycle, from new concepts to solution delivery and operations optimization.
Real-time 3D simulation is an essential tool in the development of robots and autonomous vehicles. However, content is a significant challenge in adopting 3D simulation at scale. Creating high-quality virtual worlds can be difficult and costly, requiring specialized skills, expensive toolsets, and time-consuming work. In this session, learn how the Unreal Engine from Epic Games is used to democratize simulation. Learn about the new AWS Ambit Scenario Designer for Unreal Engine 4 (Ambit), a suite of tools designed to streamline 3D content creation at scale for autonomous vehicle simulations, robotics simulations, and other real-time 3D applications.
Robotics innovation is moving at a rapid pace. That’s why it’s critical for robotics companies to accelerate and scale their hardware and software development. However, it’s challenging for robotics startups to get the time, headcount, and resources to compete with larger robotics firms with more resources and established infrastructure. For this reason, AWS created the AWS Robotics Startup Accelerator to help these companies move at the pace of innovation and get products to market quickly. This session shares results and learnings from the AWS Robotics Startup Accelerator pilot program.
Do you ever wonder how Amazon builds intelligent robots? In this session, learn about different AI-enabled fulfillment process paths that Amazon works on. Explore difficult AI problem statements and how Amazon Robotics approaches solutions to them.
Miso Robotics is bringing restaurants into the future with an array of robotics, computer vision, and AI-activated products. This session explains how Miso is leveraging services like AWS RoboMaker for new robotics applications like Flippy, the robotic fry cook. Every restaurant is different, and with the cloud-based simulation that AWS RoboMaker provides, Miso keeps its fleet of robotic cooks working 24/7. Learn how AWS RoboMaker simulation allows Miso to run a battery of simulations on every install before deploying new software to the fleet of robot fry cooks.
Seafloor Systems is reinventing how hydrographers learn about bodies of water. Hydrographic research is used to create nautical charts that help people navigate oceans, lakes, and rivers. Today, many bodies of water are uncharted, and it can be difficult and dangerous for people to collect data. In this session, learn how Seafloor’s fleet of more than 3,000 autonomous robots is helping scale underwater exploration and make new discoveries. Join to hear Seafloor explain how they built robots with the mission of making robotics more accessible to researchers, and hear about the scientific advancements made possible by Seafloor’s robots.
As our population lives longer through better science, we’ll need to develop technologies that empower seniors and other individuals to live more independently. In this session, Mike Dooley, CEO of Labrador Systems, introduces Labrador’s new Alexa-enabled smart service robot that opens up possibilities for assisting people in the home and beyond.
Reducing ML inference time is crucial for safety-critical applications like autonomous driving, satellite navigation, and last-mile delivery robots. This session leads you through the application of AWS Wavelength—our 5G edge computing infrastructure—to cloud robotics. Learn how near-real-time collision avoidance using vision data can be deployed to robots with zero fleet hardware updates. This session includes a focus on a prototype built for Bell Canada’s 5G network. See how the ability to conduct ML inference at the edge with AWS (without hosting models on individual devices) helps with more scalable, efficient, and cost-effective field operations.
Functional safety certification for autonomous mobile robots (AMRs) is highly ambiguous because standards are being written while the robotics industry is developing products and creating the process to certify them. An AMR process has been defined and is being matured that provides a path for selection of a standard, and definition of analyses that must be performed at various phases of product development to verify that requirements are met. In this session, learn how this process has been used for three Amazon products that are paving the way for a certification of autonomous mobile robots now and in the future.
The design of Amazon’s fulfilment centers has always been customer-driven, with automation solutions increasing safety, shortening delivery times, increasing capacity, and maintaining cost efficiency. With the introduction of one-day and same-day delivery promises, the scale and complexity of the network has grown dramatically, and consequently, the energy footprint is increasingly important. Amazon’s pledge to achieve net zero carbon emission by 2040 will require contributions from all sectors of our operations. This session provides an overview of the opportunities for reducing energy consumption in robotics and automation.
The Boston Dynamics Spot robot is often deployed in dangerous or unsafe environments to perform a variety of complex tasks like maintenance, operational monitoring, and construction. To perform these tasks well, robots need the ability to sense the world around them and react to unexpected conditions. In this session, Boston Dynamics demonstrates how they build and deploy ML models to their Spot robot to operate intelligently at the edge. Learn about a new capability that enables Boston Dynamics customers to train ML models tailored to their use cases.
Computer vision (CV) algorithms process and analyze visual data to detect objects and classify images in ways that mimic the human mind but at exponentially greater speed and scale. In this session, learn about the science of computer vision with an emphasis on creating high volumes of labeled images for training CV models. Discover how 3D models can be used to create virtual environments representing real-world scenarios, and how to support data security, privacy, and compliance requirements while working with the training data.
Autonomous robots must be tested in millions of scenarios so they can reliably react to changing conditions in their environment. Simulation is key to scaling testing scenarios with real-world conditions for robotics development. AWS RoboMaker has added new features, including GPU support, which helps developers run high-fidelity simulations that closely mimic the real world. In this chalk talk, learn how to build and run simulations in AWS RoboMaker using high-fidelity modern simulation engines. See how to connect a simulation engine to AWS RoboMaker, create simulation environments, and run key tests.
What if the usual functions of a café were done automatically? AWS implemented a self-order kiosk using a VR/AR cloud service, and automated the cycle of making and serving coffee through simulated robotic operations. In this chalk talk, learn about the AWS tech behind our Robotics Café. Learn how AWS RoboMaker, IoT, digital twins, and more make the Robotics Café possible.
Where are all the robots that are supposed to make our lives easier? It turns out that getting a robot to do useful things in the real world is harder than just giving your computer wheels and a camera. In this chalk talk, consider some of the surprising things that are hard for robots to do and dig into why they’re trickier than they seem. Take a closer look at when advancements in automation and AI can be applied to robots and when they can’t. Finally, explore tools that can help you get started in the world of robotics.
Autonomous mobile robots operating in warehouses need to complete tasks alongside people, pallets, and other obstacles in an always-changing environment. In this session, learn how Amazon's Canvas robotics team leverages AI for computer vision to turn imperfect sensor inputs into a semantic understanding of the warehouse, allowing for fluid and intuitive robot operations.
Robotics is growing rapidly, and simulation is an important tool for developing and testing robots. Open 3D Engine (O3DE) is a modular, open-source, cross-platform 3D engine built to power anything from games to high-fidelity simulation. ROS 2 is the most widely used open-source robot development software in the world today. AWS RoboMaker allows robot software developers to run simulations at scale in the cloud. All three together are a powerful robotics development toolset. This chalk talk provides an overview of each of these tools and demonstrates how to use them together to accelerate robotics software development.
Building autonomous robots requires a multi-disciplinary team with specialized skills in nascent technology domains such as AI/ML and simulation. AWS robotics offerings provide a specialized set of tools and solutions that make it easier and faster for robotics developers to build and operate cloud-enhanced autonomous robots. In this chalk talk, learn how to manage and deploy software at the edge, build an MLOps pipeline, and ingest and stream live video for robot surveillance and interaction.
Robots are edge devices that need the ability to collect, analyze, and store a large amount of data in real time. In this chalk talk, learn how to architect a robotics data pipeline using AWS services such as AWS IoT Greengrass and AWS IoT SiteWise. Learn how to connect to robots, reliably collect data while maximizing bandwidth, and how to set up analytics so you can take action on the collected data.
In this chalk talk, explore a day in the life of a roboticist/ML expert at Amazon Robotics. Walk through how the team prototypes ML and robotics projects from concept to production with example projects like Robin perception.
Do you want to see in augmented reality (AR) how your robot application will perform in the real world before you finish your development? AR robotics visualization on AWS displays and manages robots and robot fleets using AWS IoT Greengrass, and enables interaction in AR so you can see your interactions on an actual robot. In this workshop, learn how to use AR to interact with a virtual robot and see a robot react in both the virtual and physical worlds.
The Indy Autonomous Challenge powered by Cisco (IAC) is a $1.5 million, DARPA-style Grand Challenge Competition to develop a fully autonomous Indy-style vehicle. The teams competed in two events: at the Indianapolis Motor Speedway in November 2021 and during CES 2022 in Las Vegas, Nevada. While IAC ramps up for 2022 and 2023 races, AWS is working with them to provide a comprehensive simulation environment based on AWS RoboMaker, SVL Simulator, and Autoware’s autonomy stack to allow the teams globally to work on their software and prepare for the competition. In this workshop, walk through an overview of this environment and have the opportunity to experiment with it.
Simulation is essential to designing and developing robust robots quickly and cost effectively. Virtual world environments like NVIDIA Isaac Sim take simulation into a new dimension with high fidelity and the ability to generate a myriad of variations, yielding breakthrough training results. Setting up virtual world simulation environments can be challenging, requiring lengthy, error-prone installation scripts, tweaking, and debugging for each target platform. In this workshop, learn how to deploy and run an Isaac Sim simulation in the cloud via a Docker container with AWS RoboMaker. Learn how to quickly and painlessly set up your robotics simulations in the cloud.
An exploration of the possible (and necessary). Could robots be natural and simple enough to interact with that you interact with them like you do humans? This session explores the potential of and need for assistive technologies with real-world use cases in action. Learn the benefits of AWS RoboMaker and hear stories from AWS Partners who are focused on using robots to redefine the possible of assistive technologies.
Robotic arms are used to automate the transfer of objects from one place to another in a facility. As companies scale autonomous operations, robotic arms must be able to operate in highly dynamic environments and adjust in real time to unexpected changes. Enhancing robotic arms with machine learning (ML) capacities improves the accuracy of dynamic pick-and-place operations and reduces downtime due to unexpected changes. In this workshop, get hands-on instructions to build, test, and train AI-activated robotic arms. Learn how to test and train ML models in simulation and then deploy software to robotic arms.
The physics of traveling in space imposes challenges on real-time video communications, but with modern networking solutions and smart video collaboration technology, Webex has devised a way to keep space travelers connected. Before NASA's Artemis I mission was announced to fuel the new lunar economy, the aerospace industry couldn't fathom a future with on-demand video due to legacy thinking associated with connectivity limitations. In this session, Jono Luk, VP of Product Management at Webex, showcases the power AI and other emerging technologies can have on propelling next-generation video compression, adaptation, and resiliency technologies to revolutionize human connections in space.
The space economy is booming, with 115,000 new satellites projected to enter space by 2030. As space becomes more congested, the risk of crashes increases between satellites that are critical to our daily routines, like GPS, television, internet, and more. In this session, leading female executive and founder in the space industry Melanie Stricklan discusses how Slingshot Aerospace is accelerating space sustainability to create a safer, more connected world. Explore how Slingshot’s next-generation technologies are improving transparency, safety, and sustainability in orbit to ensure space sustains our global economy for generations to come.
Join this session to learn how Digantara is building a maps database of space using a space-based surveillance platform with global real-time Earth coverage. They are deploying a constellation of cost-efficient nanosatellites in low Earth orbit along with precise modeling through an AI/ML-based algorithm to provide predictive and robust space-based situational awareness services. Their goal is to create the world’s biggest catalog of man-made space objects. Explore how the Digantara satellites are designed to use a combination of Lidar sensors to scan other constellations to find, measure, and report on currently undetected space debris that could pose an invisible threat to all in-orbit satellites.
Companies have typically relied on expensive satellite equipment to directly deliver media and broadcast content. In an industry first, SES is using AWS to deliver content with IP connectivity via satellite directly on AWS, allowing end customers to do all of their production and processing in the cloud. In this session, SES shares how they are saving their customers money and providing innovative ways of content distribution by harnessing the global scale of AWS.
Geopolitical, economic, and environmental monitoring from space is entering an exciting era: a new generation of AI-driven satellites is watching every corner of the globe, providing previously hidden insights to governments, industry, the media, and citizens. In this session, learn how AWS-powered AI and machine learning is being applied to real-time mission management for swarms of autonomous satellites, improving transparency and accuracy. See novel uses of satellite remote sensing and geospatial analytics during crises like oil spills, supply chain impacts, and global conflicts.
The ability to access space is opening new business opportunities including commerce, research, and tourism. Orbital Reef is a premier mixed-use space station intended to launch into low Earth orbit and be available for anyone to lease access to the space environment. Whether goals are scientific research, exploration system development, manufacturing of unique products, media, or unique hospitality, customers can find a berth there. In this session, learn how Orbital Reef is using Amazon and AWS products and services to create an Earth-like experience with world-class technical accommodations and a scalable system architecture that will allow any nation, agency, culture, or individual to participate in the space economy.
Data has gravity. The data generated by permanent outposts and colonies on the Moon and Mars will be too large, time sensitive, and mission critical to send back to Earth for processing, generating a need for large-scale compute and storage infrastructure on location. This localization would create lunar and planetary enclaves that interface with Earth and each other but operate independently. During this session, consider the challenges of powering this infrastructure, producing cost- and resource-effective hardware for these environments, and building the communication and networking capabilities needed. In the process, help answer the question everyone is asking: “How will astronauts stream Netflix on Mars?”
Growth in the space industry has spurred exponential growth in space-generated data and new approaches to space-focused software applications. Traditionally, onboard applications are fixed for the mission duration, requiring vehicle operators to store and downlink collected data regardless of value, which increases latency for data-driven actions. Agility in application development and deployment provides mission flexibility and drives a shorter timeline for identifying valuable space data. Join this session to hear how artificial intelligence (AI), machine learning, and powerful processing hardware applied to space vehicles can optimize vehicle storage, activate low-latency decisions, and drive space vehicle autonomy from low Earth orbit to the Moon.
SpaceIL started with three engineers who met at a bar with dreams of building and sending the first private spacecraft to the Moon. The spacecraft Beresheet, Hebrew for genesis, was the smallest and most cost-effective spacecraft ever to reach the Moon, where it crash-landed into the lunar surface. Over one million students contributed to the project, inspiring an entire country and millions globally. In this session, Kfir Damari, co-founder of SpaceIL, shares the incredible story of how they successfully built the spacecraft, the challenges they faced, the technical details behind the crash, lessons learned for future missions, and insight into the new journey ahead.
BlackSky is leading a new era in real-time geospatial intelligence by incorporating low latency, multidomain sensing, and advanced AI into a seamless user experience. BlackSky is building multipath communication networks and edge-hosted AI to update mission plans and analyze data onboard their satellites to understand and react to observations as they happen and rapidly feed data to Earth. In this session, learn how BlackSky’s advancements and multidomain sensing provide real-time geospatial intelligence to track global supply chain dynamics, humanitarian crises, geopolitical conflict, and the impacts of climate change, all from a web-browser.
Ever wonder what it was like to grow up with an aerospace industry pioneer who contributed to the world’s most innovative communications including overseeing the design and launch of more than 170 satellites? Robert E. Berry, Sr., supported high-speed internet connections, direct-to-home TV broadcasting, audio broadcasting to cars, mobile user connections, meteorology, defense communications and air traffic control. In this session, Renee Berry shares stories and lessons she learned at the kitchen table with her grandfather, and discusses how they shaped her passion for mentoring women, working with early-stage founders, and her work with AWS.
What kind of tech will propel and sustain human environments in space? How close are we to this future? Join Lockheed Martin to look into the future of space—not generations from now but what’s around the corner. Explore a vision for the early years of lunar colonies, advancements on the ground, and new protections for space infrastructure. Underpinning that progress is machine learning, automation, and robotics, both on Earth and in space. The leader of Lockheed Martin’s space innovation labs illustrates how they’re preparing now for a future where humanity lives and thrives beyond our home world. This presentation is brought to you by Lockheed Martin, an AWS Partner.
Viasat in partnership with AWS deployed a resilient tactical edge capability where applications, like real-time image recognition on a live video feed, that require reliable, high-bandwidth communication and high-powered computing can be used in war by fighters in combat. The joint capability fused different connectivity pipes (dual-band Ku/Ka SATCOM, our LTE, and local RF net) to enable high-bandwidth capability out to the tactical edge with AWS Snowball Edge.
In this chalk talk, learn about Amazon’s Project Kuiper, an initiative to increase global broadband access through a constellation of 3,236 satellites in low Earth orbit. The system is anticipated to deliver fast, affordable broadband to unserved and underserved communities around the world. Amazon believes the internet is a net positive for users and communities, but there are still billions of people on Earth without reliable broadband access. Project Kuiper invites users to explore the art of the possible when you are no longer constrained by connectivity, bandwidth, or latency issues.
In this chalk talk, consider how entanglement activates a number of applications such as quantum computing, quantum networking, and quantum key distribution. Discuss how, in the future, entanglement will be distributed between remote quantum devices using fiber, free-space, and satellite links with the help of quantum repeaters and networks. These quantum networks will allow remote access to quantum computers while keeping their data and programs private. Quantum networks also activate quantum key distribution: the generation of encryption keys that are invulnerable to decoding by quantum computers.
In this workshop, learn how cities can plan their roadmap to achieve carbon neutrality using satellite imagery and AI to estimate the amount of carbon dioxide that urban green infrastructure can capture. Dive deep into how Latitudo 40’s cloud-based model empowers urban planners to learn from the past, monitor the present, and shape the future to achieve carbon neutrality and improve sustainability and quality of life on our planet.
In this workshop, become a space explorer by using machine learning (ML) to find exoplanets using data collected by NASA’s Kepler space telescope. Space enthusiasts only need a laptop—no prior ML experience is required—to identify planets light-years away from Earth. Using Amazon SageMaker Canvas, discover the power of AWS ML services and, hopefully, a brand new planet.
Advance your cloud adoption journey by solving challenges designed to teach AWS best practices around security, migration, DevOps, AI/ML, and more, all tailored to AWS for Aerospace and Satellite use cases. This workshop aims to help you solve challenges like considering how to stop a threat actor from moving laterally through your virtual software integration laboratory or injecting false images into your AI/ML image pipeline. If you are new to AWS or have never used a specific AWS service, this workshop helps prepare you to navigate these challenges. AWS experts will be on hand as coaches, even as the workshop promotes self-paced discovery and learning.
In space, food resources must be optimized. Traditionally, ML models base this optimization on nutrition factors and astronaut choice. However, once in space, the taste of the chosen meals often changes. Piquancy is the only flavor that is factored outside of these models to compensate for body fluid movement in microgravity. In this chalk talk, explore a new approach to modeling food inspired by flavor wheels used by chocolatiers and fragrance pyramids used by perfumers. This three-dimensional flavor model helps provide inputs to build resource-optimized food profiles. These profiles can be used to generate minimal resource-based food recipes to substitute astronaut choices.
In 2020, NASA commissioned Axiom to provide a commercial module to the International Space Station to develop the world’s first commercial space station. In this chalk talk, hear about Axiom’s plans to provide its customers with the ability to make groundbreaking discoveries, develop on-orbit capabilities for in-space manufacturing, and develop technology for deep space exploration. Learn how edge data processing is critical for Axiom’s realization of these goals.
SkyWatch is on a mission to democratize remote sensing data through a simple user experience. They built an efficient and scalable processing system for satellite data using AWS at the core of their infrastructure. Join this workshop to learn how SkyWatch built its satellite imagery solution using AWS Lambda and Amazon EFS. SkyWatch develops new applications such as SkyWatch TerraStream and SkyWatch EarthCache Earth observation imagery platforms for this data every week, and demand is increasing across many industries, including infrastructure monitoring, construction, and finance, in addition to humanitarian causes such as coordinating disaster assessment and relief.