Unabiz adds AI to boost Wi-Fi positioning on Sigfox IoT trackers

Sigfox-parent Unabiz reckons it can now deliver a 90 percent “success rate” with Wi-Fi positioning on Sigfox trackers thanks to the addition of some AI/ML trickery to its Wi-Fi scanning software. Sigfox operator KYOCERA Communication Systems (KCCS) has claimed a jump from 62 percent to 92 percent in its Wi-Fi geolocation hit rate on Sigfox-based trackers in Japan.

Singapore-based Unabiz said its upgraded Atlas WiFi service will be rolled out immediately and progressively, one network at a time, across all its 70 territories where Sigfox infrastructure is available. It said traditional Wi-Fi geolocation, achieved by passive scanning and triangulation of local Wi-Fi access points, struggled to find a positioning target up to 30 percent of the time. 

“[The] success rate… typically falls below 70 percent in some regions,” it said. It suggested the effort to raise the bar much higher has been complex and costly, notably in large-scale asset tracking projects – until now. Unabiz said it has “activated” a built-in machine learning (ML) module in Sigfox trackers to get the figure above 90 percent.

The achievement was described as “exceptional”, by the company itself, leading it to rob the old ‘de-facto standard’ marketing terminology (repeated ad infinitum by the LoRaWAN Alliance) to lay a claim to the LPWAN crown for low-power positioning. “This… reinforces Unabiz’s commitment to… the most energy and resource-efficient LPWAN solutions for the supply-chain industry,” it said. 

Alexis Susset, chief technology officer at UnaBiz, said: “The widespread coverage of Sigfox… enables geolocation data network effects which, combined with AI, has enabled us to increase our positioning success rates to levels beyond 90 percent in all regions we have tested. This… further establishes [Sigfox] as the de facto standard for energy-efficient, high-accuracy asset tracking.”

Unabiz said the upgrade is available without “additional resources or new infrastructure deployment[s]”. It added: “Instead of switching to higher-cost hardware alternatives like GPS trackers, which entail significant expenses, customers can continue to rely on their existing hardware devices to enjoy this more precise AI-enhanced WiFi-based tracking solution.”

Naoki Kawai, division manager in the wireless solutions division at KCCS, commented: “Since the adoption of the AI-enhanced Sigfox Atlas Wifi ML service, our geolocation success rate has increased… to 92 percent. We’re excited to promote this new service to our customers and look forward to continuing our strategic partnership with Unabiz in Japan’s supply chain sector.”

Henri Bong, chief executive at Unabiz, said: “Driven by the principle [that] less-is-more, this upgrade allows customers to enjoy… a higher geolocation success rate without having to switch to more expensive… and resource-intensive alternatives. [It]… reduces operational costs and… environmental impact[s] associated with manufacturing and deploying new infrastructure.”