Latency matters. That’s why it makes sense to develop computational capacity at the edge
“There’s so much data in the edge environment,” Rob High, CTO of IBM Watson told PC Mag recently, that “it makes sense to bring some of the cloud computing capabilities into the computational capacity of the edge device.”
He makes a good point. By placing the processing and AI right next to IoT devices, Edge Computers can make real-time decisions, operating with a speed, intelligence and overall reliability that just isn’t always possible with cloud-based solutions.
It’s a game-changing approach and the latest System-on-Chip (SoC) neural processors – like those in the VIA ALTA DS 3 Edge AI System – have the power to accelerate key ‘deep learning’ applications, including image identification, speech recognition, data analytics and natural language processing.
As Gartner’s Tom Bittman wrote in the article The Edge Will Eat The Cloud: “As people need to interact with their digitally-assisted realities in real-time, waiting on a data center miles (or many miles) away isn’t going to work. Latency matters. I’m here, right now, and I’m gone in seconds. Put up the right appealing advertising before I look away, point out the store that I’ve been looking for as I drive, let me know that a colleague is heading my way, help my self-driving car avoid other cars through a busy intersection. And do it now.”
Edge AI in Retail
Retail is set to benefit massively from Edge AI hardware. With systems such as the VIA Smart Retail Engagement System, stores can spot blacklisted shoppers immediately, while age and gender detection can collect data that gives retailers more insight into their shoppers.
Armed with this demographic data, digital signage can show targeted ads, while facial recognition technology can be used to auto-detect VIP customers, letting staff welcome them immediately. OCBC Bank has reported great success with its automatic VIP programme.
Edge AI promises to give shops a more intelligent way of dealing with stock and store layout too. JD.com in Indonesia, for example, has a cashier-free store that uses cameras and edge systems to help optimise inventory, digital signage and store management.
Edge AI for Transport
Fully autonomous vehicles are the future of urban transportation and Edge computing (with AI processing) will be key to their development. Using sophisticated sensors and cameras, self-driving vehicles can be enhanced with technology that makes them more efficient and safer.
Using the VIA Mobile360 ADAS system, for example, the Enchi Auto company built a self-driving electric bus, which was unveiled at the company’s manufacturing plant in Huzhou in Zhejiang Province in China.
The ability to capture data and make driving decisions in real-time is crucial. Here, road-scanning LIDAR sensors feed data into the VIA Mobile360 system, which provides a 360-degree view around the bus from four high-resolution cameras. Additional driver tools include lane departure and collision warnings, blind spot detection and speed limit detection. Cloud computing alone can’t process all this data and deliver the decision-making speed required.
Edge AI and Health
In the healthcare space, Edge Computing and Edge AI are poised to improve network connectivity and boost telemedicine/remote monitoring applications.
Wearable devices provide a basic form of Edge Computing. Watches and fitness bands can monitor the health of the wearer, processing the data on-device before sharing it over the cloud. Intelligence can also be applied. For example, continuous glucose monitoring systems can let wearers monitor their levels in real-time, allowing them to receive alerts before problems occur.
In the future, edge processing will be vital in the development of robotic surgeons. Just as self-driving cars can’t rely on a network connection or suffer the slight delays associated with cloud connectivity, tomorrow’s robo-doctors will demand latency-free control. Deploying AI at the edge (rather than in a distant data center), might mean the difference between life and death.
Edge AI and Manufacturing
Edge Computing and AI also have a big role to play in manufacturing, using local sensors to control and manage output, dramatically improving efficiency and reducing errors. The advantage here is that edge systems can respond to input in milliseconds, either making adjustments to fix an issue or shutting down a production line to prevent serious problems.
Edge AI can also introduce predictive maintenance. That is, sensors can be used to monitor key systems, with AI used to predict when an error is likely to occur. Intel, for example, uses accelerometers in the Fan Filtration Units (FFU) installed at its semiconductor production facilities. This allows it to send automatic alerts when faults are detected, create a proactive maintenance schedule and reduce unscheduled downtime.
Edge AI and You
Edge AI can respond quickly to the input from sensors and IoT devices in practically any business, letting you make real-time, business-led decisions. Better insight, increased efficiency and reduced downtime are just some of the benefits that edge AI can bring. Find out more about the VIA ALTA DS 3 Edge AI System and see what it can do for you.