Penguin Applies AI, ML to BLE 5.1 For Precise RTLS

Published: August 14, 2024
  • The location technology company has built an algorithm layer to its BLE 5.1 solution so that hospitals and other companies can gain 3-foot-based location accuracy without paying a high cost.
  • The system adds intelligence to Angle of Arrive and signal strength data to precisely identify where an off-the-shelf BLE 5.1 tag is located.

Despite a growing demand for real-time location (RTLS) systems to improve asset management and workflow at hospitals, adoption has remained limited. The high cost of RTLS infrastructure, or the limited granularity of location data from legacy Bluetooth Low Energy (BLE) or Wi-Fi-only solutions, have provided challenges for those seeking answers to the question of where their equipment and people are.

Emerging technology company Penguin Location Services is a relatively new company that offers what it considers a low-cost alternative to RTLS that is accurate and high performance.

The company’s solution leverages BLE’s latest version (5.1), but with an additional AI and machine-learning algorithm layer to identify sub-meter level granularity for asset- or person-tracking. The solution is now piloted in several industries, including healthcare and oil and gas.

BLE 5.1 Plus Intelligence Layer

The BLE 5.1 version enables usage of antenna arrays and a Multiple Signal Classification (MUSIC) algorithm for angle of arrival (AoA) based estimation. Multiple antennas within a locator device can help calculate the angle at which transmission was sent.

That still poses some shortcomings though, said Mohammed Smadi, Penguin CEO, as reflectors in the environment can lead to Bluetooth transmission from multiple angles at once.

As a result, Penguin developed algorithms to help make sense of the transmission data. The system is designed to consider the signals across multiple antennas, even in environments with RF reflection.

“Our methodology improves accuracy and reliability,” said Smadi.

Leveraging a History in Navigation

Penguin, launched in 2018, initially concentrated on indoor navigation—providing blue dot-based positioning of people’s badges or smart-devices, with maps and location-based messaging and some analytics.

Following market research, the company has expanded into technology-based tracking to include RTLS solutions for asset management in healthcare. In fact, the company found that despite the need for RTLS based data, only about 15 to 20 percent of North American hospitals have an RTLS solution in use.

“When we dug deeper, we realized that cost was the major factor in adoption,” Smadi explained.

Penguin’s Concentration

In fact, he said, many early adoptions stalled or have not expanded as expected because of some basic challenges. Wi-Fi systems require tags that consume a large amount of power, thereby requiring frequent, costly and time-consuming battery replacements. Ultra-wideband can be expensive to deploy as well. Although it provides highly-precise location data, the tag devices are also power-hungry, and an infrastructure of receivers is also required.

While BLE offers a lower cost solution, the current versions don’t provide enough reliable location data to calculate the kind of room level accuracy that healthcare organizations demand.

So Penguin’s efforts centered around developing algorithms that leverage the growing amount of computational horsepower that is available now, both on the edge and in the cloud, Smadi said. That includes advancements in AI both in the edge nodes and in the cloud.

The AI algorithms the company has developed leverage environmental and behavioral data that it learns over time, to overcome any RF reflection and to correct location misestimates. That can mean correcting data related to reflection in a specific space.

Keeping Cost Down

The solution works with standard, off-the-shelf BLE 5.1 tags and locator beacons. These tags have become affordable, often several dollars per beacon tag.

When it comes to the software, “we split the algorithm in such a way to optimize for cost,” said Smadi. That means employing cloud environments, segregating data and enabling what’s called statistical gains.

Typically, users only pay a monthly subscription, specifically related to the number of assets being managed, and when the system is used to track them.

“If you look at the total cost of ownership of a given RTLS [solution] today you’re looking at the tag, you’re looking at the location engine, you’re looking at the anchor nodes, you’re looking at the cabling but there is also an insignificant cost associated with the RTLS application layer,” he said.

Easy Solutions

In contrast, Penguin’s solution serves as an out-of-box application layer while tags can be procured for most BLE vendors. Penguin is selling its BLE 5.1 development kit using off the shelf hardware and a sample software application layer as well as its AI-based location algorithm.

The company is taking advantage of the fact that more Wi-Fi nodes are being built with BLE 5.1 functionality built in as well, so that some hospitals or other companies would not need to deploy dedicated locator infrastructure.

Currently users would need to add beacon devices in addition to the Wi-Fi nodes, however, to achieve the needed location granularity. That may change in the future as the technology gets more precise.

“Our objective is to move into a lower density of anchors with a goal of having the same density as Wi-Fi access points that are [already] deployed,” stated Smadi.

ZulaFly Finds High Level Data Accuracy

RTLS company ZulaFly is working with Penguin’s solution to provide accurate location data for its customers.  That accuracy is essential, said Stephanie Andersen, ZulaFly’s managing partner. “When the location data is extremely accurate, end users have faith in the system.”

In fact, an RTLS system that doesn’t work properly can lead to inaccurate reporting and can cause “alert fatigue,” from excess reporting, Andersen said. ZulaFly relies on hardware partners to provide accurate data so that we can collectively provide a reliable and dependable solution to our customers.

As a solution provider, the Fargo, ND-based company integrates with a number of disparate systems and multiple technologies to provide the best solution for each of its customers use cases. To that end, ZulaFly thoroughly vets and tests each hardware technology that we partner and go to market with everyone’s best interest.

“We have found the Penguin technology to be highly accurate,” said Andersen. “It is important to keep in mind that not all BLE technologies are the same and specifically not all BLE 5.1 technologies are not the same.  As the platform that collects the location data, we have first-hand experience and have seen the difference.”

Learn More:

About the Author: Claire Swedberg