Over the next decade, the complexity and load placed upon enterprise networks will rapidly outpace our ability to manage them effectively. This will constrain business growth, create performance bottlenecks, and undermine the experiences network teams are working to deliver.
The leading indicators of this crisis are easy to measure. In the Cisco 2021 Global Networking Trends Report, a survey of over 600 enterprise networks found an average of 4,400 monthly events impacting wireless networks. As organizations rush to enable a hybrid workforce, we expect to see that flood of issues rapidly increase.
For hybrid organizations, connectivity is a core network service, and interruptions caused by authentication, DHCP, DNS, and other errors can’t be allowed to grind the business to a halt. But when IT spends the majority of its time on troubleshooting hundreds of events a day, it’s impossible to find time to do the important work needed to drive business growth. The capacity of the team—and the network they support—suffers as a result.
The coming storm: preparing for the Great Resignation
Compounding these capacity challenges is the looming threat of the Great Resignation, the fear that a newly empowered remote workforce will pick up sticks and seek greener career grass. In another 2021 survey, Enterprise Strategy Group found that 55% of IT leaders indicated that their biggest skill gaps were in network architecture, administration, and planning. That’s a capacity problem that’s very hard to remediate.
To break this bottleneck and avoid the single points of failure that threaten network resilience, IT teams need better tools to help them adapt faster, troubleshoot with more precision, and free up time to properly plan for future capacity needs.
Putting network data to work for customers
At Cisco Meraki, we see a future where intelligent, data-defined networks give CIOs the power to readily adapt to change, effectively increase capacity, and find better ways to plan for the future. This is at the heart of a data-defined network, an architecture that not only can quantify the impact of an event, but provides precise evidence about the root cause and quickly points toward a recommended remediation.
There are three key elements that can unlock these kinds of data-driven improvements:
Using data to drive outcomes. Going beyond basic analytics and insight, any investment in AI and ML must come in the service of a desirable customer outcome.
Depth and breadth of data. When it comes to making relevant, contextual recommendations, more data beats a better model every day of the week. The best platforms can learn from data captured not just from your network, but from millions of other networks like yours. That breadth and depth is critical to informing effective decisions.
Upholding trust. Having the data and the models is not sufficient. A data-defined platform must overdeliver on its ability to protect and govern the data it captures, continually improve the quality of its recommendations, and always keep control in the customers’ hands. Think autonomous, not 100% automated.
We’ll be talking more about our outcome-focused approach to AI and ML systems at Meraki in the coming months. In the meantime, we’d love to hear from you in the Meraki Community about which network challenges are impacting your team’s capacity and how more intelligent services might help define the future of your network.
In keeping with the age-old Meraki philosophy of empowering our customers to do more with less, people-counting analytics on MV12 has finally arrived!
We’re excited to see this new set of tools build on top of an already impressive (and necessary) security product. Now, MV12 can act not only as a great security camera, but also as a sensor for businesses big and small — no servers or extra infrastructure needed.
If you’re already familiar with our wireless product line, this rollout might feel reminiscent of our WLAN Location Analytics tool, and it should! At Meraki, we love the notion of providing our customers more intelligence with less infrastructure, an idea especially apparent with MR and now MV.
How does it work?
Using the advanced processor on our recently launched MV12 security camera, and built-in, anonymized person detection (not to be confused with unique facial recognition/identification) software, video is stored and analyzed on-camera, at the edge. This metadata is sent to the cloud and aggregated into people-counting metrics independently of the video itself. Plus, over time this software will become more accurate using machine learning.
To see this functionality in action, just click on the ‘Analytics’ tab for an individual camera and select the time resolution (minute-by-minute, hourly, or daily) and timeframe of interest. The ‘People count’ section of this page shows an at-a-glance overview of your busiest time period, estimated peak occupancy over that period, and the total number of entrances. Remember that the ‘Total Entrances’ value will double count individuals if they leave a frame and then return, since this data is anonymized. Consequently, we encourage thoughtful placement of cameras intended for use as sensors to minimize both double counting (place them in an area with restricted traffic flow moving in one direction, like an ‘Entrance Only’ door) and occlusions (where two people or objects pass in front of one another, making it difficult for the camera to see what’s going on).
Clicking on the ‘Most Utilized’ and ‘Peak Occupancy’ results will jump directly to that moment in the camera’s historical footage so you can quickly analyze what events may have driven that spike in traffic. Drilling down into each bar in the people counting bar chart will also take you to the corresponding piece of footage, making it simple to investigate anomalies.
You’ll now be able to observe and quantify granular foot traffic patterns through a given frame.
For retailers: monitor the ebb and flow of customers throughout the day, optimize staffing headcount to make sure your customers get the attention they need, and increase the efficacy of marketing campaigns by targeting days of the week with the greatest or least traffic.
In schools: track general attendance patterns, see which areas of campus are used most frequently, and make a business case for updating facilities and equipment based on usage patterns.
At offices: figure out whether it makes sense to add more common spaces, or repurpose these areas based on popularity with office-dwellers. And are those pricey espresso machines actually getting used anyway?
Of course, these examples represent only a fraction of the uses cases now available with this additional functionality. Coupled with motion heat maps (available on all MV models), it’s never been quite so easy to see the big picture quickly.
Does this mean my MR Location Analytics setup is now redundant?
Definitely not! Think of these tools as complementary. Because MR access points count mobile device wireless signals throughout a wireless network, they provide a broad “macro” level view of foot traffic through, say, an entire store location. People counting on MV only tabulates traffic within that visual frame, making it more accurate on a “micro” level, like an individual product display within that store. By pairing these two features, you can quickly gain insights across multiple levels of your business.