Conferences and supporting programme
Scaling IoT with edge computing and heterogenous Arm devices
In the last few years we have seen an exponential increase in the number of IoT connected devices being deployed in the field. The ability to quickly connect devices to the cloud has made it easier to rely on it for any computational requirements. Even though this approach has advantages, such as low cost of deployed devices and easier control, it brings additional complications to the overall system.
The volume of data to transfer from endpoints to the cloud can be very high, the latency of the response can vary considerably and the connection can be interrupted. Depending on the application, these issues could effectively exclude the cloud as an option for data processing. As an example, a car can’t rely on the cloud to process the images from a camera to avoid an obstacle.
An emerging solution to this problem is to move some of the processing from the cloud back to the edge nodes: this trend is known as edge computing. Decentralized decision-making is also one of the design principles promoted by the Industry 4.0 initiative aimed at creating smart factories.
We will discuss edge computing as it relates to device security, reliability, and performance. In this talk we will explore how heterogeneous multiprocessor (HMP) devices can fulfil the requirements of edge computing by providing the real-time capabilities of a typical IoT node together with a powerful application processor able to perform processing and analysis directly on the device.
We will demonstrate an effective way to allocate tasks on a heterogeneous device and on the cloud and the best way to improve communication between them. We will discuss a reference software architecture for HMP IoT systems. Finally, we will look at power management and how to maximize the battery life of IoT devices.
--- Date: 27.02.2018 Time: 12:00 PM - 12:30 PM Location: Exhibitor's Forum, hall 4, stand 4-428