IOTech Webcast Series
Join Andrew Foster, IOTech Product Director and Tom Urquhart, PhD., IOTech SVP for Global Solutions in this webinar to explore the challenges of Time-Critical computing at the IoT Edge - how platform technology can accelerate development and deployment of maintainable and portable solutions and the types of use cases where this technology is applicable. Plus... watch Edge XRT in action! - a demonstration of the first Time-Critical IoT Platform for the Edge.
- What are the characteristics and challenges faced by Time-Critical edge systems
- The importance of edge platform technology to meet these challenges
- What are some of key Time-Critical use cases at the edge
- What capabilities do the edge software platforms need to provide to support Time-Critical use cases
- What’s the business case for buy versus build
- Edge XRT Time-Critical Edge Platform in action and the numbers
There is a class of IoT application at the Edge that requires very fast response times (microseconds vs. milliseconds). In many cases, these applications also need to deliver guaranteed or predictable responses to support the time-critical automation systems using high-frequency data for control or analytics.
More often than not the devices these applications run on will be severely resource-constrained (e.g. embedded controller), with available memory measured in a few hundred Kbytes vs. Mbytes.
Until recently user choice has been limited to proprietary or bespoke solutions to support time-critical application needs at the OT edge. These solutions have been based on dedicated hardware (e.g. PLC or industrial controller) with limited software programmability, making them inflexible as a system evolves.
As well as being able to upgrade ‘brown field hardware’ with more flexible software infrastructure users also want the ability to deploy on commodity versus proprietary hardware.
The current generation of IoT Edge Platforms don’t really help, being designed to run on general purpose IoT Gateways and on-premise servers, supporting data processing rates measured in hundreds of milliseconds to seconds and certainly not offering ‘predictable performance’. Until now!