As companies deliver increasingly impressive tech-based experiences, consumer and B2B expectations are rising. “Smart” is the new bare minimum; to be a leader, you need to deliver instant intelligence both within and beyond your company’s walls.
In Trend 5 of the Accenture 2018 Technology Vision, we explain the technological backbone that companies must build to meet these expectations the “Internet of Thinking.”
“Edge” devices are growing in numbers & power. How will business #ITinfrastructure adapt? Details on @accenturelabs’ solution in a #TechVision2018 blog: www.accenture.com/smartsolutions
Delivering intelligence everywhere means moving more computing power to the edge, where interactions happen, and balancing the power of the cloud to deliver meta-insights and long-term improvements to edge algorithms.
Click here to read“Find Your Edge: Bringing analytics to data at the edge of "IoT," Accenture" Labs' PoV on edge analytics.”Accenture Labs’ PoV on edge analytics.
It’s a new evolution in enterprise infrastructure, and also a new IT wrinkle: Companies will introduce more powerful edge devices into their networks, but they’ll still need to maintain legacy IoT devices and networks alongside their newer, more intelligent counterparts.
What’s more, new edge devices capable of onboard analytics deliver intelligence in previously untouched environments, where power or connectivity constraints prevented IoT deployments.
But most companies’ infrastructures aren’t set up to support this spectrum of applications and devices yet. That’s an IT headache of epic proportions—so how will companies adapt?
Accenture Labs has developed a solution to this challenge, using a knowledge graph to maintain the necessary information about each device in the network, as well as the needs and capabilities of the infrastructure.
When hearing “knowledge graph,” most people think of the framework that search engines like Google use to deliver comprehensive results. But more generally speaking, a knowledge graph is a structure that categorizes different pieces of data by the way they relate to each other.
A knowledge graph would link concepts and item types via relationships: “Mona Lisa” is “art,” “created” by “da Vinci,” and so on. It would also link the year it was painted, and other information related to the painting.
In a retail supply chain, a knowledge graph might link inventory information with the product supplier, order numbers, routes taken during its shipment between warehouses, and so on. This application lets companies systematically store expert knowledge in a graph formation—one from which an artificial intelligence solution can infer and deduce information.
Eventually, by combining natural language processing with AI and knowledge graphs, companies will be able to query their knowledge bases with a standard question format. That includes relatively simple questions like “Who painted the Mona Lisa?” but also extends to more contextual queries like “Why is inventory low?”
The Accenture Labs Systems and Platforms R&D group is using knowledge graphs to help clients in manufacturing, as well as oil and gas, maintain new devices alongside legacy IoT devices. The knowledge graph solution collects critical information about each device in the network, from its manufacturer, to what operating system it runs on, what protocols it can understand, whether it has constant or occasional connectivity, and more.
This domain knowledge graph gives the clients a powerful new way to manage their increasingly complex infrastructures, while also laying the groundwork to answer questions like “Which brand of devices do we usually use on ocean rigs?” or “How long would we need to provide temporary connectivity at an offline location to collect data that’s been gathered from the devices there?”
This kind of approach is becoming a critical need as clients build the Internet of Thinking, trying to extend their situational awareness across varied physical environments.
“Clients found that they had more and more things with different protocols trying to connect to their system,” explains Colin Puri, Technology Research Principal at Accenture Labs. “This new approach helps manage all that information and stores the knowledge gained from installing these devices, so you can go from generation to generation of technicians without loss of knowledge.
“When you come back with a new or similar device, you can automatically make a recommendation on how to deploy it, and so on,” explains Puri.
As companies look to build the Internet of Thinking—even in environments that were previously out of reach—this knowledge graph approach to device management provides a powerful advantage. It handles the heavy lifting of device management and institutional knowledge transfer, letting companies focus on what they really want to deliver: intelligence.
For more information about the Accenture Labs Systems & Platforms group, visit our website. To request a demo or information session, contact Teresa Tung at email@example.com.
Find out more about the Internet of Thinking trend and read the full 2018 Technology Vision at https://www.accenture.com/technologyvision.