Building Up the ‘Edge’ for a Responsible Society

Convenience is just one benefit of edge computing. Beyond that, it can help us build a better society, one that’s more socially responsible by reducing energy consumption, keeping us safer, protecting our data privacy and improving productivity. In this episode of the Smarter World Podcast, Nitin Dahad, editor at embedded.com and EE Times, discusses the topic with NXP’s Gowri Chindalore, head of edge strategy and Amanda McGregor, head of product innovation for applications processors.

We’ve come a long way since we converted mechanical devices into smart devices and connected them to enable the Internet of everything. With processing prowess, the edge was created for computing to occur closer to where the data is created in the physical world. It’s here at the edge where devices can respond incredibly fast, without being slowed by latency introduced by faraway cloud processing. And now, with advancements in machine learning and inference capability in edge processors, we have the intelligent edge, which can offer massive efficiency gains.

The anticipated tens of billions of connected devices must be designed to use as little energy as possible, Chindalore says. A 5-megabyte image from your smart doorlock sent to the cloud to be processed 20 times a day is thousands of times more energy-expensive than if the image is processed locally on the device. But for intelligent edge devices to be truly efficient you must investigate how the device is designed to use energy. You should be able to turn power on and off in different parts of the devices where only the active part of the device gets power. The device can remain in a deep sleep, low-power state but still be alert of its environment. It can always listen for a wake word, for example, but power on the full system only after prompted.

The next evolution is the aware edge where intelligent devices connect and collaborate with each other across a network, sharing insights and interacting. This awareness is the key to unlock real energy efficiencies, safety and productivity.

-Gowri Chindalore

Edge technology also unlocks tremendous opportunities to improve safety in our homes and at work. McGregor points out a unique example of edge processors embedded in smart hardhats that process sensor data in real-time to provide industrial workers crucial insights on heat, oxygen levels, fatigue, toxic conditions and other stresses for safety on the job. The data collected from sensors on the hats and in the environment can be processed in real-time at the edge without a network connection.

There’s also increasing concern about protecting consumer data and personal privacy. The edge helps address those concerns because it processes the data where it is generated rather than sending it to some remote location where it could be vulnerable to attacks. It can be taken one step further by adding machine learning and intelligence to the edge so that it not only processes the data but also makes a decision based on the information that it just collected, making sure that all the data, processing and information is stored within that local system. That's where a trustable system comes into play, Chindalore says.

He also explains how the next evolution is the aware edge where intelligent devices connect and collaborate with each other across a network, sharing insights and interacting. These devices are context-aware and they communicate with each other to make meaningful decisions. This awareness is the key to unlock real energy efficiencies, safety and productivity.

We're seeing real-world examples of the aware edge being put into play now. The opportunities are accelerating because we're helping build the technology that can bring this vision to reality.

-Amanda McGregor

We're seeing real-world examples of the aware edge being put into play now—McGregor adds. The opportunities are accelerating because we're helping build the technology that can bring this vision to reality. It comes with challenges, though, such as how we deploy efficient ML on devices, as many of the models still need to be trained in the cloud. They must be optimized and deployed to run at the edge.

To make the aware edge a reality, it's vital that the ecosystem, regulations and standardizations are built where different products from different companies and different players can interact with guiding principles.

Listen to the full podcast, Building Up The ‘Edge’ for a Responsible Society with host Nitin Dahad, editor at embedded.com and EE Times, NXP’s Gowri Chindalore, head of edge strategy and Amanda McGregor, head of product innovation for applications processors.

You can subscribe to the podcast on Spotify and iTunes.