Virtual Iot & Edge Days 2023

At its core, Edge Computing distributes computational power and data storage wherever needed. From a developer's perspective, the whole point is to bring cloud-native tools and techniques outside the data center. Naturally, this is useful in a wide range of AI use cases. After all, bringing the model closer to the action reduces bandwidth usage and latency, making AI more effective.

If you leverage AI in the Cloud, you probably deploy your workloads on Kubernetes. But is Kubernetes needed on the Edge? Many open source and commercial Edge platforms currently available orchestrate applications and services in their own way. Others, on the other hand, integrate with Kubernetes. Which one is the right approach? In this session, you will learn how to answer this question from an AI perspective through examples related to industrial automation and connected vehicles.

Have you ever struggled to integrate embedded devices in your application logic? Especially when your application may also run everywhere across the cloud to thing continuum? For a long time, stitching together multiple communication technologies has been the obvious way to go for making microcontrollers part of your application. But a new time has come.

Eclipse Zenoh has proven to be a unified solution for the Cloud-to-Things continuum for efficiently distributing data, greatly improving the wire efficiency of existing protocols while providing higher level abstractions.

In this talk, you will learn about Zenoh-Pico: a Zenoh-based implementation that helps developers to support constrained devices in their application logic. You will also learn how commonly used protocols compare with Zenoh-Pico with respect to performance, latency, footprint, wire overhead, and energy consumption. And finally, the fun part: you will learn how to build an embedded application based on Zenoh-Pico to remotely control a tiny and cute robot across the entire Cloud-to-Things continuum.

Remember the LAMP stack (Linux, Apache, MySQL and PHP) from the late 90’s when the World Wide Web was exploding into the public consciousness? As you’ll recall, LAMP came about as an answer to a set of challenges that the web development community were facing. At that moment, Open Source came to the rescue!

Today, our team works predominantly within the Embedded Systems and Internet of Things (IoT) domain, but many of the challenges we face are parallels to those faced by the creators of the LAMP stack. This is why we think that the open source MING (MQTT, InfluxDB, NodeRED and Grafana) stack can enable the next generation of IoT Application developers.

With this masterclass we aim to take attendees through the journey we’ve been on as we’ve been developing Internet of Things solutions.

We talk through how Open Source solutions exist to fulfill these needs today, and how we arrived at the MING stack:

• Mosquitto for MQTT based assured data-comms using a publish/subscribe model
• InfluxDB for time-series data storage
• Node-Red for intuitive wiring of “nodes” to ingest telemetry data, transform, and deliver internally or to 3rd Parties
• Grafana for data-visualisation and alerting

There are a set of core generic requirements to most if not all of these types of IoT projects: Assured data-communications, Data storage, Business Logic, Visualisation

We will cover how each of these Open Source technologies can be managed and deployed using container platforms.

Internet of Things (IoT) device vendors must deal with several constraints: limited memory, processing power, and the feature list getting longer daily. It might seem challenging to bring security into the mix, and developers sometimes add security features as an afterthought. Is it possible to improve IoT security without too much effort? In this talk, Marta will show you how to match security with the improved general quality of your product. She will walk through the audience of her favorite five steps, starting with removing unnecessary features. With examples. You can apply those techniques to most IoT operating systems and software stacks.

MQTT has emerged as one of the dominant IoT pub/sub technologies due in large part because you can publish basically any payload on any topic to fit the application. But as MQTT matures into vertical industries within the "Industrial Internet of Things", the MQTT Topic Namespace and Payload contents need to be well defined for device and application interoperability.

The Sparkplug specification defines an Industrial centric Topic Namespace and an efficient Payload definition targeting industrial process values such as temperature, pressure, level, rpm, etc.. The Sparkplug specification also provides well defined MQTT session management as well as well defined MQTT STATE management.

This presentation will provide an overview of "why" the Sparkplug specification was developed and how it is used in real world industrial IoT solutions.