By Len Schuch, GM of 5G Positioning Services, ASOCS
The rise of industrial and enterprise positioning
Life without GPS would seem almost impossible today. Reliance on navigation apps has become part of our daily lives and when added to features such as “Find My Phone”, Apple Tags, and a plethora of other applications, it seems we are all looking to find things. Where GPS excels in outdoor environments, alternative Real-Time Tracking Systems (RTLS) like Wi-Fi, Bluetooth, RFID, UWB, etc., have emerged as solutions for indoor tracking. However, most of these solutions necessitate the installation of dedicated network equipment for positioning, posing challenges in areas that encompass both indoor and outdoor spaces as well as increasing the overall costs of the solution. Think about automated factories, smart warehouses, hospitals, airports, and the list goes on.
Even outdoors GPS has its limitations, with accuracy typically reaching up to a 4.9-meter radius, GPS is not secure, takes time for an initial fix, and can be unreliable in a dense building environment. The problems do not end there, in real-life settings there are obstacles, moving parts, and interference to take into account.
5G Positioning can provide the necessary accuracy for over 90% of use cases, does not need any additional equipment over and above what was deployed for the original 5G Private Network, is 5G native, and therefore does not require any additional hardware on the device, reducing costs both in terms of installation and maintenance.
A few companies have begun to demonstrate 5G Positioning and there have been some good results in clean, uncluttered lab conditions and we welcome the showcasing of the importance of 5G Positioning for the industry.
Here at ASOCS, we decided to do things differently and have been looking at getting results in non-sterile environments that mimic real-world manufacturing settings.
Real-world requirements for positioning
The digitization of Industry is a driving force for deploying 5G Private Networks and is key to increased process automation and optimization. Moreover, knowing where assets, tools, equipment, and even people have a direct impact on increasing efficiency and generating cost savings.
Positioning is the foundational application for 5G Private Networks. Across industry verticals: smart manufacturing and warehouses, hospitals, airports, ports or mines, etc. with each sector having a multitude of different use cases from tracking AMRs and on-site vehicles, to locating assets, connected tools, and providing geofencing safety zones for employees.
The list of use cases and applications across commercial and industrial settings meets the needs to accurately locate but have live positioning for any number of devices.
Private 5G – the foundation for providing positioning
The inherent properties of Private 5G (P5G) key have all the key components and infrastructure needed for positioning, which meets the requirements for over 90% of industrial use cases.
Mobility
The most obvious property is mobility – a cable-free wireless infrastructure with performance to match cable. Once Radio Units (RUs) have been installed, adding devices is as simple as connecting them wirelessly to the network.
Uniform Coverage
Unlike Wi-Fi which offers “data hotspots” that require handover with lag time between Wi-Fi coverage zones, Private 5G offers single, seamless, and uniform coverage. Need to increase the coverage area? Add more Radio Units and rather than create new zones, you simply increase the coverage area of the network. As simple as it sounds, the key to accurate positioning is dependent on a reliable single coverage network.
5G Positioning only requires a minimum of three RUs and not necessarily in Line of Sight (LoS) but what makes it more interesting is that the radius for location coverage is four times the size of the 5G coverage from the same network infrastructure, making it very adaptable for even the most challenging environments.
Guaranteed SLA
Private 5G provides guaranteed and reliable connectivity. Devices are tiered meaning that there is full control over which ones get priority and bandwidth. With Private Network Slicing there is also full control over which devices are connected to URLLC and which on eMBB depending on the data and speed required.
Security
As only the devices you register can connect to the network, combined with the inherent 3GPP protocols, security, and privacy are at the core of a Private 5G network.
For positioning, all of the location calculation is conducted on the network or cloud, nothing resides on the device itself.
AI – adapting to real-world settings
In sterile lab conditions with numerous radio units, good positioning results can be obtained. Real life does not happen in a laboratory; in the real world, there are obstacles, radio interference, and moving parts.
The answer to overcoming these difficulties is Artificial Intelligence (AI), and an ongoing learning process to compensate for real-time discrepancies while reducing the number of radio units (RU) needed. Not only are fewer RUs needed, but each RU can also be utilized for tracking beyond the network coverage area.
In live testing in an underground parking lot, with cars and pillars, three radio units reached initial accuracy of under 2 meters (compared to 4.9 meters from GPS). This number declines with each iteration as the AI learns to better understand and adjust for interference.
Even at 2 meters, in a 10,000-to-50,000-meter industrial plant or campus, this is more than enough to find equipment such as forklifts, industrial parts, and materials. The expectation is that the AI can reduce this to sub-meter to decimeter accuracy.
Positioning to become the norm
Just like the wider adoption of GPS, AI-driven Positioning over Private 5G is set to be not only the norm, but essential for operations in a wide variety of commercial, enterprise, and industrial settings.
Although in its infancy as of 2023, AI is growing up fast, while the need for indoor and campus positioning services increases daily. The future is coming much faster than anyone can predict.
