Machine Learning a crucial part of network management, according to survey

Network professionals are increasingly convinced that machine learning is a crucial part of network management, according to a study by network analysis company Kentik. According to the research, this conviction partly stems from the increasing use of multi-cloud environments.

To find out what network professionals think about using AI, 388 network experts participated in a Kentik survey at the Cisco Live US conference in San Diego last month. Half of these participants work as network technicians, the other half as management staff, software architects and developers. The participants represented a diverse selection of fields, including education, energy, finance, government, healthcare and technology.

ZDNet reports that 65 percent of participants said they thought machine learning was very important for network management. That is an increase compared to the 45 percent from a comparable study by Kentik in 2018. The company combines the interest in machine learning with the growing use of cloud environments. As many as 76 percent of participants say they use cloud services, of which almost half (47 percent) use a multi-cloud strategy.

“People and manual processes can no longer keep pace with network innovation, evolution, complexity and change,” says Jim Frey, VP strategic alliances at Kentik, in a statement. “That’s why we hear more about self-driving networks, self-healing networks, being intent-based networks and other concepts.”

Energy sector the most, health care the least prepared

Of the survey participants, people from the energy sector showed the most interest in machine learning: 75 percent called it “extremely important” or “very important” for network management. No person in this group considered machine learning “not important at all”. On the other hand, respondents from the health sector showed the least interest, with only 41% calling machine learning ‘extremely important’ or ‘very important’.

With the exception of the technology industry, it also emerged that the energy sector is the most prepared for full automation. 30 percent of respondents from the energy sector indicated that their organisation is “extremely well prepared” or “very well prepared” for full automation. In the healthcare sector, only 3 percent of respondents indicated that their organisation is “very well prepared”.

The study also showed that automation is often used for the configuration of networks, with 53 percent of the respondents citing this as their intended use. Policy management was the second most automated process. This was mentioned by 40 percent of the respondents.