Real-time meets real world: Use case discussions
It’s one thing to talk about AI in telecom. It’s another to make it work in the real world. In this panel, operators, technology partners, and infrastructure providers moved beyond theory to share how AI and data strategies are actually being applied in the field — from IoT platforms and smart cities to autonomous drones and real-time orchestration.
The takeaway? AI success isn’t just about models. It’s about how you collect, connect, and use data across complex, multi-partner environments.
IoT Is Where Some of the Most Valuable Data Lives
While telecom has traditionally focused on subscriber data, the panel made a strong case for a shift: some of the most valuable data today is coming from IoT.
From sensors in buildings and vehicles to connected infrastructure in cities, IoT is generating massive volumes of real-world, real-time data. But that data is:
- Highly distributed
- Owned by different entities
- Often siloed across systems and platforms
The challenge and opportunity is to unify this data into something usable.
By aggregating data from multiple networks and devices into low-code platforms and dashboards, operators can:
- Monitor environments in real time (temperature, presence, water, etc.)
- Enable new enterprise services (smart buildings, parking, logistics)
- Begin exploring data monetization opportunities
But raw data alone isn’t enough. The real value comes when that data is processed, contextualized, and made accessible — often with AI layered on top.
From Devices to Decisions: AI Needs Streaming Data
One of the examples came from our TM Forum Catalyst project, focused on IoT-driven drone services.
In this use case:
- Thousands of drones provide security monitoring for ports
- Real-time telemetry streams data on location, performance, and energy usage
- AI models optimize operations and sustainability
- Telecom systems dynamically adjust 5G network slices based on demand
- Usage is tracked and monetized in real time
What makes this possible isn’t just connectivity — it’s architecture.
By using event-driven architecture (EDA), all of this data — from business systems to device telemetry — is streamed, orchestrated, and made actionable in real time.
This enables:
- Instant decision-making (e.g. upgrading video streams to 4K)
- Automated orchestration across systems
- Real-time assurance and failure detection
- New monetization models tied to usage
It’s a clear example of how telecom can move from static systems to dynamic, data-driven services.
AI in Action: From Smart Cities to Safer Roads
Beyond telecom-specific use cases, the panel highlighted how these same data principles apply across industries.
In one example, a collaboration between Dell, AT&T, and the city of Bellevue used AI and real-time data from traffic cameras to:
- Reduce congestion
- Predict unsafe pedestrian behavior
- Dynamically adjust traffic signals
- Improve public safety
This required processing terabytes of data daily, analyzing it in real time, and feeding it into AI systems that could act immediately.
It’s a powerful reminder: When data is accessible and actionable, AI can move from experimentation to real-world impact — fast.
Speed Is the New Requirement
A consistent theme across all speakers was speed.
The old model — 18 to 24 month transformation cycles — doesn’t hold up when:
- AI use cases demand real-time data
- Business value depends on rapid deployment
- Outcomes impact safety, revenue, or customer experience
So what’s changing?
1. Break the Problem Down
Start small. Focus on specific use cases. Deliver value quickly, then expand.
2. Make Data Easy to Access
Use connectors, low-code tools, and streaming architectures to simplify how data is ingested and shared.
3. Design for Consumption
Data should be delivered in the format and flow that downstream systems — including AI — can immediately use.
4. Prioritize Outcomes Over Perfection
Don’t build for the sake of technology. Build for results.
Ecosystems, Not Silos
None of these use cases were built in isolation.
Every example involved multiple partners:
- Telecom operators
- Infrastructure providers
- Software vendors
- AI specialists
- Enterprises and public sector organizations
This reflects a broader shift in telecom: Speed now depends on collaboration.
Operators are no longer building everything themselves. Instead, they are orchestrating ecosystems — combining capabilities to deliver outcomes faster.
Rethinking Data: From Internal Asset to Shared Value
One of the most forward-looking ideas from the panel was around data sharing.
Today, much of IoT data remains locked within individual organizations. But there is growing recognition that:
- Sharing anonymized data can unlock new services
- Cross-industry collaboration can improve outcomes (e.g. safety, sustainability)
- Data can be monetized beyond its original use case
Real-world examples already show this potential:
- Sensor and location data used in disaster response
- Vehicle data used to improve driver safety
- Infrastructure data used to optimize cities
This isn’t just about better business outcomes. It’s about better societal outcomes.
The Bottom Line: Data Is Only Valuable If You Use It
Across every example, one message stood out:
Collecting data isn’t enough.
If data is:
- Sitting in silos
- Difficult to access
- Not used to drive decisions
…it has no value.
The operators and enterprises seeing real results are the ones who:
- Make data accessible
- Apply AI in context
- Turn insights into action
- Build systems that scale
Because in the end, AI doesn’t create value on its own. Data — connected, contextual, and actionable — does.
