AI at the edge: Driving new business opportunities at the point of data creation

05 May, 2020
Stephen Meserve

Data has always come from outside the data center. As the number of smart devices generating new data grows, there is a need for faster processing and an increased pressure on networks to process data efficiently. The processing must necessarily leave the data center and move closer to the edge, where the data are created. By 2022, 50 percent of enterprise data will be processed at the edge, compared to only 10 percent today.

Edge computing enables faster insights and actions as running inference closer to the data source improves speed to action. In AI inferencing, where sub-second latency is the expectation, the latency accumulated by sending data back to a central processing facility becomes the bottleneck in business operations.

With dedicated infrastructure, better data security and control are possible, as minimizing data transport to central hubs reduces both vulnerabilities and the aforementioned latency. Finally, putting processing at the edge enables continuous operations to reduce disruption and cost, running autonomously even when disconnected.

When the creators of the Mayflower Autonomous Ship faced these exact challenges, they reached out to IBM to deliver a solution that would meet their difficult requirements. The team at ProMare was looking to create a fully autonomous version of the Mayflower ship to recreate the historic Atlantic Ocean crossing in 1620 from Plymouth, England, to America. Sailing across the Atlantic Ocean represents the ultimate edge application, thousands of miles from any human contact or data center.

The team decided to put IBM at the helm, building an AI captain using IBM Power Systems AC922 and IBM Visual Insights. Without human intervention, the ship will navigate the Atlantic after having been trained using a literal ocean of images to recognize and maneuver past anything it encounters on the way.

“Many of today’s autonomous ships are really just automated robots [that] do not dynamically adapt to new situations and rely heavily on operator override,” says Don Scott, CTO of the Mayflower Autonomous Ship. “Using an integrated suite of IBM’s AI, cloud, and edge technologies, we are aiming to give the Mayflower full autonomy and are pushing the boundaries of what’s currently possible.”

You can learn more about the Mayflower Autonomous Ship here.

We also recently announced IBM Power Systems IC922, built specifically for inference workloads at the edge. Power Systems IC922 provides the compute-intensive and low-latency infrastructure needed to unlock business insights from trained AI models. POWER9-based, it supports faster data throughput and decreased latency through advanced interconnects such as PCIe Gen4 and OpenCAPI. Both are important for edge AI applications.

As 5G rollout continues across the world, edge computing combined with 5G creates opportunities to enhance digital experiences, improve performance and data security, and enable continuous operations in every industry. Edge brings computation and data storage closer to wherever data is created by people, places and things. We can provide an autonomous management offering that addresses the scale, variability and rate of change in edge environments, in edge-enabled industry solutions and services, and in the solutions that can help you modernize your networks and deliver new services at the edge.

Edge computing with IBM is ready to help you go from data to insight today. Learn more now.

The post AI at the edge: Driving new business opportunities at the point of data creation appeared first on IBM IT Infrastructure Blog.