How Context Turns Location Data into Operational Intelligence

For years, location data has been one of the most widely discussed applications of IoT. From asset tracking to workforce visibility, organisations have invested heavily in technologies capable of pinpointing where people, equipment and materials are located.

Yet despite advances in Real-Time Location Systems (RTLS) and indoor positioning, many deployments still struggle to deliver meaningful business value.

The reason, argues Samuel Van de Velde, Founder and CTO of Pozyx, is that the industry has been solving the wrong problem. Knowing where something is only tells part of the story. The real opportunity lies in understanding why it’s there, what it’s doing, and what that means for the business.

As Samuel explains in the most recent episode of the IoT & AI Leaders Podcast, accurate location data becomes transformative when it is enriched with context. Combined with Industrial AI, organisations can move beyond simple tracking to understand workflows, automate decisions and build increasingly intelligent operations.

Listen to the full episode here: https://www.iotleaders.ai/podcast/beyond-the-blue-dot-how-ai-gives-location-data-meaning/

Why location data alone isn’t enough

Indoor positioning has existed for years. Modern RTLS technologies can locate assets with remarkable precision, often within centimetres, but the technology itself isn’t the challenge.

The challenge is what organisations do with the information afterwards. As Samuel explains, much of today’s tracking still resembles a digital version of ‘Find My Device’. A blue dot on a map showing where something happens to be. That may be useful when searching for missing equipment, but it rarely delivers the operational improvements businesses expect from their IoT investments.

Samuel argues that value can be truly seen once more contextual questions are being asked. How is an asset being used? How much time does it spend idle? Which workflows create bottlenecks? Where are delays consistently occurring? How efficiently are people and equipment working together?

From tracking movement to understanding behaviour

Samuel explains that Pozyx was tasked with locating cattle across large farms, using accurate indoor positioning to allow behavioural patterns to emerge.

By identifying how long cows spent eating, resting, drinking or visiting different parts of the barn, AI models could infer much more meaningful insights about health and breeding readiness. The raw X and Y coordinates themselves weren’t valuable; the behavioural patterns derived from them were.

This illustrates an important principle, where location data becomes significantly more valuable once it acquires semantic meaning. Rather than collecting coordinates, organisations begin collecting evidence of behaviour.

The same principle extends far beyond agriculture. Whether monitoring manufacturing operations, logistics facilities or industrial assets, understanding how objects move is often far more valuable than simply knowing where they are.

What is location intelligence?

Location intelligence is the process of transforming raw location data into meaningful operational context.

Rather than analysing coordinates in isolation, organisations interpret movement against real-world environments, workflows and business processes.

For example, a location system may recognise that an asset has:

• Entered a production workstation
• Remained idle for thirty minutes
• Completed a logistics journey
• Reached a storage area
• Been moved significantly more frequently than expected

Each of these represents a meaningful business event rather than a stream of positioning coordinates.

This contextual understanding enables organisations to ask much richer operational questions while dramatically reducing the amount of raw data people need to interpret. As Samuel explains, geo-fences, movement history and activity summaries provide AI with information it can reason about far more effectively than continuous streams of positional data.

Why manufacturing stands to benefit most

The manufacturing sector demonstrates perhaps the clearest opportunity for contextual location intelligence.

Many factories still rely on barcode scanning to track work orders, materials and inventory as they move through production. While effective, these systems depend on manual intervention. If a scan were to be missed, the ERP or manufacturing execution system would no longer reflect reality.

By contrast, RTLS enables work orders, materials and equipment to be tracked automatically as they move through production, creating a continuously updated picture of how work is progressing across the factory floor. This unlocks several operational benefits, such as a clearer understanding of work order progression, material movement, labour utilisation, equipment usage and more.

Automation becomes possible when context is understood

Context, therefore, clearly enables automation.

Samuel describes a manufacturing environment where incoming work orders automatically trigger logistics requests as materials approach production stations. Instead of relying on manual communication, the system understands where work is progressing and initiates downstream activities automatically.

Equally important, manufacturers gain visibility into whether materials are on their way, reducing costly production delays caused by missing inventory.

Why AI needs context, not raw location data

As AI becomes increasingly central to industrial operations, many organisations assume the answer is simply to collect more data, but Samuel argues the opposite. Large Language Models excel at reasoning over language. They are far less effective when presented with enormous streams of raw sensor readings or positional coordinates. Feeding millions of location updates directly into AI creates unnecessary complexity, increased cost and limited business value.

Instead, AI performs best when information has already been transformed into meaningful operational events.

Rather than analysing thousands of positional updates, AI might receive insights such as:

• Forklift completed twelve journeys
• Equipment remained idle for four hours
• Materials arrived late to Workstation Three
• Operator spent eighteen minutes assembling a product

This semantic layer dramatically reduces complexity while allowing AI to reason about workflows, identify anomalies and answer operational questions much more effectively.

The emergence of an operational intelligence layer

One of the most compelling ideas discussed in the episode is the emergence of what could be described as an operational intelligence layer.

Instead of treating IoT systems as repositories of sensor data, organisations begin building systems that understand activities, workflows and relationships. A layer sitting between raw IoT devices and AI. Rather than replacing existing enterprise software, it complements it by providing a far more accurate representation of what’s happening in the real world.

Could AI generate factory software on demand?

Looking to the future, Samuel suggests that AI could fundamentally change how operational software is created.

Today, organisations often rely on expensive development projects to build dashboards, interfaces and specialist applications. Future users may simply describe the information they need, with AI then generating lightweight, task-specific applications tailored to that exact workflow. Rather than one large enterprise platform trying to satisfy every requirement, organisations could deploy numerous small applications designed around individual operational tasks.

Samuel shares one practical example, whereby a customer wanted a simple visitor registration system that linked visitors to tracking tags while they moved around a factory. Instead of commissioning an entirely separate software solution, AI enabled the rapid creation of a lightweight kiosk application that solved the specific requirement with minimal development effort.

The future of AIoT is contextual intelligence

IoT data is one of the richest sources of operational insight available to organisations, but only when it’s interpreted within the context of real business processes.

As manufacturers continue investing in AIoT, location intelligence and Industrial AI, competitive advantage will increasingly come from understanding behaviour rather than merely recording movement.