Nyansa's Voyance uses machine learning to improve Wi-Fi troubleshooting

Wi-Fi woes trip up company business, but a tool that uses machine learning can help. Read this review that covers one company's experience with Nyansa's Voyance tool.

For all the innovation in the network management industry, traditional vendors have not exactly invested in Wi-Fi troubleshooting. In a 2017 survey of 100 North American enterprises with over 1,000 employees, ZK Research explored the issue of Wi-Fi troubleshooting and uncovered some startling results. Below are some of the findings from the survey, which highlight how big the problem has become:

  • Forty-seven percent reported at least 10% of users have weekly Wi-Fi problems.
  • Sixty percent spent more than 25% of their time troubleshooting Wi-Fi issues.
  • Fifty-nine percent said packet capture is the most commonly used tool.
  • Sixty-nine percent said the process of identifying the problem is much longer than fixing the issue.

In terms of the survey's findings, CIOs should worry about the amount of Wi-Fi problems, because workers rely heavily on Wi-Fi to get their jobs done.

The survey also revealed the majority of network engineers dedicate more than one day a week to fixing Wi-Fi issues. Considering the cost of network professionals, it's easy to see Wi-Fi problems are costing companies huge amounts of money in worker productivity and IT time.

Anyone who has worked on a network knows packet capture is done when no other options are available; it's often considered the tool of last resort. This explains why so many engineers spend so much time on Wi-Fi troubleshooting.

Moving forward, Wi-Fi is expected to connect many IoT devices. Consequently, if companies do not get a handle on these Wi-Fi problems soon, their inability to solve Wi-Fi issues could significantly hurt their businesses.

Wi-Fi troubleshooting to boost worker productivity

Excelitas Technologies Corp., a manufacturer of photonic and optoelectronic technologies, is looking to address these kinds of Wi-Fi issues. The Waltham, Mass., company has about 6,500 employees in 19 global locations, and it services several industries. Like most businesses, Excelitas relies on Wi-Fi to support its mobile workers.

For example, employees on the manufacturing floor need to use collaboration tools, update customer relationship management records and perform other tasks from their tablets. At Excelitas, Wi-Fi problems hinder worker productivity.

According to Amit Shah, CIO at Excelitas, the tools the company was using -- traditional network management products that focused on reporting and monitoring -- could not pinpoint the root cause of the Wi-Fi issues. As a result, the IT department was often scrambling and one step behind.

Nyansa Voyance uses machine learning to improve Wi-Fi troubleshooting

Given the importance of Wi-Fi, the situation at Excelitas was no longer tolerable. Shah's team researched several vendors before picking Nyansa, a network analytics company that uses machine learning to pinpoint Wi-Fi issues.

Editor's note: Nyansa is a client of ZK Research.

The vendor initially launched its Voyance product to work with controller-based Cisco and Aruba services, which make up about 70% of existing Wi-Fi deployments. Recently, Nyansa added ExtremeMobility controllers and Aruba's Instant networks to its data sources. A review of this tool found that Nyansa is the only vendor-agnostic Wi-Fi troubleshooting tool and offers WAN troubleshooting and an endpoint client agent.

Dashboard is easy to interpret

According to Shah, Nyansa was easy to learn, presents data in an intuitive dashboard and uses machine learning to turn massive amounts of data into actionable insights.

Before using Nyansa's Wi-Fi tools, the Excelitas IT department would spend hours searching for the root cause of Wi-Fi problems and often never find them. Nyansa's Voyance dashboard can help Excelitas understand the source of an issue in under 15 minutes, Shah said. It's original tools informed Excelitas of a problem, but didn't provide any guidance on how to fix it.

An introduction to and overview of the Nyansa Voyance tool.

IT groups can also drill down on granular information provided by Nyansa's product. For example, IT can see what happens when users go from one access point to another, or from location to location. As a result, IT departments can make configuration changes to improve performance.

In addition, the Nyansa dashboard enables the help desk, rather than second-level engineers, to troubleshoot certain issues. Having high-paid engineers root out the cause of Wi-Fi issues is not a good use of company resources. Alternatively, the help desk could isolate the source of the problems and sometimes fix it without involving senior IT people, allowing them to focus on more strategic issues.

Troubleshooting Skype for Business issues

Despite the positives listed in this review, Shah did note that Nyansa could improve by finding the cause of Skype for Business calling issues. Through an API, Voyance ingests detailed, per-client call metrics from Skype for Business servers and correlates this data with the rest of the stack.

Nyansa's Voyance analytics platform also provides this same capability with Cisco's Unified Communications Manager and Ascom's voice over Wi-Fi service that uses Android-based Myco smartphones to transmit secure voice and data over Wi-Fi. Discrete syslog messages from each phone are analyzed and correlated with data from the rest of the network for greater service assurance of what has become a business-critical application and popular fixture in the healthcare market.

Administrators of Skype for Business have experienced problems with how flaky the service can run over Wi-Fi. So, understanding the source of dropped or muffled calls or poor quality is important. Given how quickly Nyansa's Voyance isolates Wi-Fi issues, Shah said he'd like it to do the same for Skype for Business. Nyansa has said this Skype for Business issue is on its roadmap.

Dig Deeper on Network management and monitoring

Unified Communications
Mobile Computing
Data Center