WiFi Now recognises Aprecomm’s AI-based architecture to be the most consequential and promising technological shifts in recent years
By Claus Hetting, Wi-Fi NOW Chairman & CEO
If you’ve been in Wi-Fi for a while you will know that the transition to AI-based architecture is likely one of the most consequential and promising technological shifts in recent years. Now Aprecomm – an India-based startup with skills honed by a Wi-Fi chipset giant – is delivering the software stack allowing any Wi-Fi hardware and hence any Wi-Fi service to become AI-powered, the company says.
What’s the value of AI-based architectures for Wi-Fi? We may have discovered (at least in part) the answer to that question a couple of weeks ago when startup Mist Systems was acquired by Juniper Networks for a staggering US$405 million. Now a new startup is promising to deliver AI-based Wi-Fi practically everywhere using a brand new hardware-agnostic approach to AI.
Aprecomm – founded in 2016 by a group of ex-Qualcomm Wi-Fi engineers – says the value of AI-driven operations and tuning is indisputable because there is (outside of AI) currently no way to intelligently correlate network settings with user experience. “We’ve developed our solution by precisely measuring the user experience and dynamically correlating this with data extracted from the network,” says Pramod Babu, co-founder of Aprecomm.
Aprecomm has named their technology the ‘Self-Organizing Device stack’ or SOD™. An SOD™ unit (meaning an AI-enabled Wi-Fi router) forms part of a distributed neural network or ‘SOD™ cluster’ of devices. Most of the intelligence of the system is provided at the edge and the clusters hence form the AI backbone. Each device gleans data from behavioural patterns and auto-adjusts to boost performance. For efficiency only limited data is shared in the Cloud, Aprecomm says.
“To develop our approach to AI we’ve asked the question: How would a wireless expert – meaning a human being – tackle this problem? One classic problem in Wi-Fi is that conditions that cause bad Wi-Fi can be difficult to recreate and hence the trouble shooting can be close to impossible. Our platform pro-actively seeks out and suggests the solution,” says Pramod Babu.
Aprecomm says the economic benefits of AI-driven Wi-Fi performance optimisation are considerable. Up to 70% of the total cost of ownership of a Wi-Fi network can be attributed to operations. “Now imagine any Wi-Fi network with any Wi-Fi hardware running the Aprecomm AI stack. We believe this will add tremendous value right across the industry,” says Pramod Babu.
For more about AI-optimised Wi-Fi don’t miss meeting Aprecomm at this year’s Wi-Fi NOW USA event in Washington DC. For tickets & more information click here.