Machine learning on resource-constrained IoT edge devices
Who is this presentation for?
- Data scientists, IoT practitioners, decision makers, IoT architects, and data architects
Uniquely identified devices over a network, popularly known as the internet of things (IoT) throws some challenges when it’s implemented in a real-world, practical scenario. As we all know, even though billions of devices are getting connected, projecting an exponential growth of IoT market over months, but equally, the security and privacy of such devices or sensor hubs are a growing concern. Besides security and privacy, the main purpose of IoT is to gain immediate insights into a situation that humans take some time to reach—hence demanding near-zero latency information. High bandwidth and continuous availability are some other factors that serve as the baseline requirement for IoT to work in a real-life, practical scenario.
Deriving insights from the data collected by IoT sensors at the edge is one of the most popular solutions adapted by various industrial use cases. Sukanya Mandal focuses on such methods and demonstrates a practical use case (borrowed from a real-world scenario) on machine learning training in the cloud and inferencing at the edge. You’ll walk through the virtuous circle of the cloud and edge that forms the essential infrastructural foundation for IoT to be effective in a practical environment.
You’ll see a demonstration using machine learning capabilities on the cloud provided by Amazon SageMaker as well as the IoT capabilities provided by the AWS IoT suite—AWS IoT Core for systems engineering and AWS Greengrass for edge computing capabilities—performed on a Raspberry Pi using open datasets captured by sensor arrays to train machine learning models on the cloud using Amazon SageMaker. Once the training is completed, it’s deployed the same on the Raspberry Pi using Amazon Greengrass and then the trained model derives inference on Raspberry Pi. This demonstration precisely elaborates on how the cloud and edge computing together play a crucial role in deriving the real ROI of IoT for various industries.
- General knowledge of IoT and machine learning
What you'll learn
- Learn how and when to implement cloud and edge solutions
- Understand the necessary architectural considerations for designing IoT solutions and that edge and the cloud together gets the job done
- Explore common considerations for architecting for data science on IoT data
Sukanya Mandal is a data scientist at Capgemini. She has extensive experience building various solutions with IoT. She enjoys most working at the intersection of IoT and data science. She also leads the PyData Mumbai and Pyladies Mumbai chapter. Besides work and community efforts, she also loves to explore new tech and pursue research. She’s published a couple of white papers with IEEE and a couple more are in the pipeline.
Leave a Comment or Question
Help us make this conference the best it can be for you. Have questions you'd like this speaker to address? Suggestions for issues that deserve extra attention? Feedback that you'd like to share with the speaker and other attendees?
Join the conversation here (requires login)
Premier Diamond Sponsors
Premier Exhibitor Plus
For conference registration information and customer service
For more information on community discounts and trade opportunities with O’Reilly conferences
For information on exhibiting or sponsoring a conference
For media/analyst press inquires