IoT and Machine Learning Help Drive Network Transformation

IoT and Machine Learning Help Drive Network Transformation

To leverage emerging technologies like IoT and machine learning in business, early adopters find they need a new approach to networking.

 

Artificial Intelligence (AI), machine learning and the internet of things (IoT) lead emerging technology conversations across the world. Companies recognise that these technologies are ready to be used to drive real business benefits.

The Asia Pacific and Japan (APJ) region is set to pick up the pace on these two fronts. According to a recent cloud survey by MIT Technology Review Custom and VMware, more than 70 percent of non-users of AI in APJ said their organisations will adopt the technology within five years. IDC forecasted global IoT spending to surpass $1 trillion USD in 2020, with APJ leading the way.

As APJ businesses ramp up investment and adoption in IoT and machine learning, what does this mean for IT and networking infrastructure?

The Road to IoT and Machine Learning

 

MENACE is an early example of IoT and machine learning, in that it gradually became better at playing noughts and crosses.At a recent emerging technologies event in Asia, Bruce Davie, CTO for APJ at VMware, talked about one of the earliest learning machines, MENACE. Designed by Donald Michie in the 1960s, the “Machine Educable Noughts and Crosses Engine” gradually learned to play tic-tac-toe more proficiently with each new game. Fast forward to today, machine learning has made significant advances. It has, for instance, outstripped humans at many image recognition tasks. A striking example is the classification of a set of images of puppies and muffins. The error rate for such an algorithm in 2010 was 30 percent, but by 2016, after enough exposure to training data, that error rate dropped below 4 percent, outperforming humans. The common theme of both examples is the way algorithms improve over time with increases in both compute power and increased exposure to data.

“The concept of machines that learn has broad applications. Many of the challenges we face today, in cybersecurity, for example, will benefit from the continuously self-refining characteristics of machine learning systems,” says Davie. Machine learning can help transform security from a process of chasing bad to ensuring good, by understanding the intended (“good”) state of an application and automatically detecting and responding to deviations from that state.

With its vast compute requirements and dependence on large amounts of data, machine learning frequently relies on the cloud. India and China are particularly bullish about adopting these emerging technology trends. In China, local tech giants Tencent, Baidu, and Alibaba received government support in their quest to develop these platforms across autonomous driving, healthcare, and smart cities.

Data Centres to Centres of Data

 

In APJ, IoT is becoming a top business focus. Seventy percent of companies view the technology as critical for the future success of their organisation. The enthusiasm for IoT is by no means unfounded. More than half of adopters in the region say IoT made significant improvements to their market competitiveness.

VMware APJ CTO Bruce Davie, IoT and machine learning, networking transformation expert.

Bruce Davie, VMware VP and CTO, APJ

At the same time, at least half of APJ is still not digitally connected. As more people and connected “things” like vehicles, factory and healthcare devices come online in the region, infrastructure and architecture significantly changes. Data is more distributed, moving from the corporate data centre to the cloud and from the cloud to the edge.

“Our apps are no longer nicely contained within the walls of our data centre. Many companies today are in a mixed environment of data centre and cloud, with some applications moving to the edge,” says Davie. “IoT drives the need to process data closer to where the action takes place.”

More than 40 percent of APJ respondents in the MIT Technology Review survey strongly agree that IoT requires clouds to play a larger role in managing massive amounts of data at the edge. As the enterprise IT environment becomes more hybrid and complex, a new kind of networking architecture is required.

A New Approach to Networking

 

With the rise of IoT and machine learning, data and applications are becoming more distributed, from data centres to clouds to edge. This distribution of data and applications creates a new challenge for businesses and IT.

“Each new architecture comes with new risks and concerns,” says Davie. “Companies are trying to figure out appropriate controls around networking and security for applications and data in the cloud. Now, apps and data are moving out to the edge. This puts IT teams in a situation where they struggle to achieve consistency in security and policy across all these environments.”

The rapid amount of change over the last 15 years is challenging the traditional network model. What new network approach will companies need to face the next 20 years or more?

“The network of the past delivered a set of capabilities tied to physical devices in the data centre. The highly distributed environment of the future, however, will require a Virtual Cloud Network,” says Davie. “Virtual cloud networking is software-based, helping businesses securely deliver the data applications they need, wherever they are located—from heart monitors in hospitals to connected cars in cities and wind turbines in rural regions. Security is built into the architecture, so businesses can securely manage data distribution from the data centre to the cloud to the edge.”

CIOs in APJ lead their global counterparts in IoT and machine learning enthusiasm. For forward-thinking businesses, these emerging technologies can spark new innovations and profit streams. To successfully unlock the true potential of emerging technologies like IoT and machine learning, businesses must embrace a Virtual Cloud Network that sharpens their edge—not one that holds it back.