The Path to Digital Won’t Be Paved with Yesterday’s Networks

TwitterLinkedInGoogle+Facebook

For innovation, better productivity, and performance, organizations must invest in modern, automated networks.

By Paul Gillin, Senior Contributor, Connected Futures


Imagine having a family member in the hospital recovering from surgery, and their blood pressure drops. An alarm should go off immediately informing the medical staff. But the alarm doesn’t go off.

That’s a scenario a Boston-area hospital believed was the right scenario. That’s because they implemented an automated system that routes notices from patient monitors to medical staffs’ mobile devices.

By reducing reliance on audible alarms, the hospital believed healthcare professionals would have more flexibility to move around the hospital while still being able to respond quickly to alerts.

But there was a flaw in their planning. The software worked perfectly. The network was a different story.

“There was so much latency in the network that it took up to four minutes to get the message to the device,” said Zeus Kerravala, principal analyst at ZK Research.

“The organization overspent at the application level, but didn’t invest in the network,” continued Kerravala. “It removed the ROI completely.”

The story underscores the criticality of networks in nearly every aspect of today’s business operations.

Everything from business models to workforce needs and customer expectations are all changing drastically. And rapidly.

Meanwhile, billions of intelligent devices are flooding our networks over the next five years.

How are CIOs expected to cope with all of this? One answer is to invest in a network that can quickly align to all the needs of the business. That’s one of the best ways to successfully manage this deluge of devices and forge ahead with break-away innovation.

Those who build a responsive network will enable their organizations to see enormous gains in productivity and innovation.

Networks can no longer be constrained by rigid hardware limitations and capacity estimates. Instead, they must anticipate the needs of people and organizations using them. And then intelligently deliver services as needed.

“Digital transformation, in one word, is speed,” said Kerravala. “The network itself needs to become a much more agile and dynamic resource in order for the company to become more agile.”

Cloud computing is another major force driving the development of these intuitive networks. Information assets and applications will increasingly reside outside of on-premises data centers, making network performance and security the baseline for business agility.

“Most of us still think of networks as utilities that connect machines and people together, but that view underestimates the network’s value as a source of business opportunity, particularly as the universe of connected devices grows,” said Kevin Bandy, Cisco Chief Digital Officer.

“When you’re talking about many things being connected, machine-to-machine communications and artificial intelligence, that’s workflow,” Bandy said. “It has a material impact on how you derive value for your work.”

That impact will only grow as organizations become more digital. By some estimates, half of the Global 2000 companies will derive the majority of their business from digitally-enhanced products, services, and experiences by 2020.

Failure to pay sufficient attention to network agility and performance can seriously undermine those transformation initiatives.

Gartner claims “The future-state network is an aspirational view of how enterprise network architectures should evolve to meet emerging business requirements and be more closely aligned with critical business objectives.” And that means that “the network must become simpler, more agile, and more automated.” 1

In a word, networks need to become more intuitive. They need to understand the people, processes, and devices that are connected to them and automatically adapt to their needs on the fly.

Automation, analytics, machine learning, and artificial intelligence will enable them to continually reshape themselves, learning as they go in order to support application service levels, security policies, and operational processes.

“The speed of business won’t tolerate manual configuration and management. Automation is essential,” said Barry Libenson, CIO at Experian. “Autonomous networks that self-monitor and self-correct not only drive efficiency and reliability, but free up critical resources to focus on other tasks and drive down operational costs. They also provide an additional level of security to help mitigate against risks the network is exposed to every day.”


A Case in Point

How will this new era of networking change the way your organization operates? Consider the example of a consumer electronics manufacturer that wants to cut its product development and production times in half.

To do this, the company must accelerate design schedules, outfit production facilities for agile manufacturing, and improve demand forecasting.

The intuitive network enables the company to adopt 24-hour “follow the sun” development using design teams in Jacksonville, Seoul, and Lisbon. As one team prepares to head home for the evening, another team is arriving at the office.

To keep the teams informed, a high-definition video-conference — or “telepresence” — meeting is activated between sites.

The daily meeting has been scheduled on a shared calendar, which is network-aware. As the meeting time approaches, the telepresence system communicates its service level, security, and priority requirements to the network.

The network responds immediately to deliver the required service and security levels.

Members who can’t make the telepresence session can attend via their mobile device. The network automatically allocates the additional bandwidth to their Wi-Fi-connected phones or laptops.

A half hour before this handoff meeting, the network sets priorities and, if necessary, allocates extra capacity to handle the rapid exchange of design files, which are each several gigabytes in size.

These large files are stored on a secure network segment that only team members can access. When the teams sit down, they both have the latest schematics to work from.

Priorities for the design teams are partially set by customers. Sensors built into the company’s electronics regularly connect with a secure, cloud-based analytics engine.

This shows which features customers are using and which ones they struggle with. There are ongoing smartphone app-based surveys, which periodically asks the customer advisory panel for feedback on product performance and potential new features.

All of this information continually streams into the company’s analytics system to help plan new products and features.

Agile manufacturing is informed by demand forecasting, which is now in real time. The company gathers data from partners, social chatter, and media coverage and runs this data through machine learning-based predictive analytics systems.

The resulting projections inform IoT-networked assembly line machines, which can be recast to accommodate forecast changes in demand. As result, delivery times are short. Outages are rare.

In the factory, the network constantly monitors and protects every IoT device and machine on the floor. The network ingests millions of simultaneous data streams and compares them against historical performance benchmarks to look for signs of impending malfunction or security-related threats.

Repair orders are dispatched before equipment fails. Threats are identified and isolated instantly. Downtime is cut to nearly zero.

The scenario described here isn’t science fiction. The technology exists today, and it can be applied by any organization. The process starts with leveraging the intelligence of an intuitive network.

 

How it Works

These modern networks are software-driven. In other words, new functionality is delivered across the network. It uses open interfaces at the application level to enable it to easily “talk” to devices and applications.

Powered by machine learning and artificial intelligence principles, the network conforms to the priorities of the business, continually learning from applications and usage patterns.

For speed and flexibility, the network provides end-to-end service quality that ensures that the appropriate service levels are always met.

It can allocate and de-allocate capacity and prioritization dynamically for any specific application or service based upon policies.

Those policies are defined by network administrators to conform to the needs of the business. For example, real-time, high-fidelity voice and video sessions are essential each day to ensure the successful handoff of work between teams.

The company’s collaboration and scheduling software is integrated with the intelligent network through application program interfaces that enable the network to take cues from events in the calendar.

Policies state that priority is to be given to the telepresence session that appears in the daily calendar. Once the policy is set, any change to the calendar triggers a corresponding change in network priorities.

Security is critical, of course. One way the intuitive network addresses this issue is by flexibly creating secure network segments. These segments can be dynamically defined and changed at any time to provide protected connectivity for a set of users, devices, resources, and services. Access and permissions are based upon identity rather than location or internet address. Permissions can be automated through policy and adjusted or revoked at any time.

Built-in analytics and artificial intelligence not only optimize capacity, but also offer usage insights that can help IT organizations operate at optimal efficiency and the business make better, timely decisions.

Customer feedback is gathered using social media monitoring tools, embellished with location analytics and sentiment analysis. This enables organizations to capture time and location data from mobile devices to understand how geography, time of day, and other usage scenarios affect opinions. Customer feedback can thus be categorized and prioritized more precisely.

To accommodate data from factory floor sensors and machines and integrate with dynamic scheduling and tooling, the network infrastructure extends to seamlessly include any type of connected device.

The factory floor uses dynamic segmentation to organize the network into multiple secure production environments. Security and quality of service policies can be applied to each segment to protect the service levels and availability of these business-critical environments. Data from the devices can also be streamed into cloud-based analytics applications for instant diagnosis and predictive maintenance. The DNA network gives the organization flexibility to change the way it views and manages the network of intelligent devices at any time.

 

1Gartner, Danilo Ciscato, Mark Fabbi, Lisa Pierce: 2017 Strategic Roadmap for Networking, February 2017, G00324263