Focus: Private 5G


Why many private 5G deployments fail before they start?


By Tamer Kadous, General Manager, XCOM RAN, Globalstar T


he private 5G market is growing fast. T ere were 6,500 private LTE/5G networks deployed globally


by the end of 2025, with spending projected to exceed $7.2bn by 2028. T e need is clearly there, but the story


coming from engineering, operations and IT teams in factories, warehouses and port yards is diff erent. Many complain of stalled pilots, budget overruns and lack of seamless connectivity. So, what must change? For a start, the industry must stop trying


to force-fi t complex architectures and products built for the telecom industry, and start using enterprise-grade solutions purpose built for industrial settings.


A telecoms playbook in an industrial environment T e gap between ambition and execution comes down to one problem that rarely gets named plainly: traditional private 5G was designed for telecom networks and not industrial environments. T e deployment models, RF planning requirements and multi-vendor integration chains were all inherited from carrier networks built to cover entire cities. Dropping that architecture into a dense, metal-heavy factory fl oor is where deployments unravel. A factory fl oor is not a blank canvas. It


already has PLCs, manufacturing execution systems, SCADA infrastructure and a patchwork of OT protocols that have been accumulating for decades. Connecting a new private 5G network to all of this requires integration work that most industrial buyers only discover aſt er the radio hardware is already installed. When that work is underestimated, the


network gets deployed but can’t actually serve the equipment it was meant to connect. Most who deploy conventional small-


Connecting a new private 5G network to legacy systems requires integration work that most industrial buyers only discover after the radio hardware is already installed


cell networks assume you can plan your way around interference: site surveys and mapping coverage zones, carefully separating radios so they don’t step on each other. T at works in relatively open environments, but not in a warehouse where the RF environment shiſt s constantly and the instinct to add more radios to get more capacity backfi res. Every additional radio just compounds the interference problem. T e result is that many manufacturers


end up with expensive proof-of-concept networks that technically function but can’t carry the workloads they were commissioned for. Mission-critical applications like autonomous mobile robots, AI-based quality inspection and real-time process control demand something closer to 99.999% uptime and sub-20ms latency. An industrial network is not a network that occasionally drops handovers or degrades under load when production peaks.


Dedicated industrial architecture T e shiſt now is away from telco-style deployment models and toward adopting


08 July/August 2026 www.electronicsworld.co.uk


solutions aligned with how enterprise IT/ OT actually operate. T e architecture runs on standard x86 hardware rather than proprietary radio units, with the O-RAN split between radio and processing stack. T is means the network evolves through soſt ware updates instead of hardware replacement cycles. Crucially, it can be managed by an IT team without a specialist RF engineer required on site. T e architecture change that matters


most is moving from independent small- cell radios to joint, centralised processing across multiple radios. Interference disappears entirely, and a device moving through the facility stays connected to the same logical cell when all radios share a single processing brain. No handovers. No disruption. A deployment can be densifi ed without


the performance degradation that kills conventional approaches, because capacity scales linearly rather than degrading as radios start competing. Cooperative radios can also overlap freely, which removes weeks of RF planning. T ey’re mounted, the system is brought up, and centralised processing handles coordination automatically, which is the diff erence between a deployable system and an aspirational one.


Spectrum strategy is part of the architecture decision Spectrum choice adds another layer of risk, particularly across Europe. Shared or contended spectrum – most notably CBRS in the US – creates exposure for mission-critical applications if access is re-prioritised or contested. Dedicated licensed spectrum, where an organisation has exclusive interference-free access across multiple countries, is materially diff erent for applications where availability is non- negotiable. T at decision belongs in the


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44