Two fairs, two different worlds, but one clear message for the embedded industry: the future belongs to those who can bridge the gap between “visionary” and “viable”. Having visited both CES 2026 in Las Vegas and Embedded World in Nuremberg, the team at intive observed two sides of the same coin.
CES once again showcased what is technically imaginable today: bold visions, impressive demonstrators and plenty of AI-powered promises. Embedded World, on the other hand, focused on what is technically feasible: real products, rigorous constraints and real engineering challenges. A look back to both events combined provides a surprising picture of where the embedded sector is heading – and where it should slow down (and think twice).
AI is everywhere, but the usefulness is often optional
Artificial intelligence has become the default label for innovation for some. At CES, AI was attached to almost everything. At Embedded World, the conversation was more grounded.
What became obvious at both fairs: AI can deliver value, but only if it solves a concrete problem. And if there is no problem, why should it be added to the system?
In the embedded sector, constraints still matter. Limited compute power of microcontrollers, energy consumption on battery operated devices, and deterministic behavior do not disappear just because an AI model is involved. Many showcased solutions were impressive demonstrations, but their added value in real-world embedded products remained questionable.
The takeaway is simple: AI is not a feature. It is a tool. And like any tool, it only makes sense when used knowingly.
Edge AI: from trend to necessity
While “AI everywhere” feels inflated, Edge AI remains a critical necessity.
The drivers are clear:
- Low latency requirements
- Data privacy and regulatory pressure
- Reduced cloud dependency
- Cost and energy efficiency
Processing data directly on the device is no longer a niche approach. It will be a key architectural decision for certain embedded systems.
What matters now is not whether Edge AI is used, but where it makes really sense, how models are deployed, and how they are maintained over long product lifecycles.
This is where embedded engineering experience becomes critical.
Humanoid robots: impressive showcases, unclear use cases
Humanoid robots were the stars of both shows: walking, talking, interacting and drawing crowds. The halls even featured live spectacles like robot boxing matches.
From a technical perspective, they are great engineering examples. However, from a product perspective, the use cases remain hard to justify regarding cost, safety, and maintainability.
Most of these humanoids felt more like technology showcases than answers to concrete industrial problems. They demonstrate advances in perception, motion control, and interaction - but for now, humanoids seem to be less about immediate adoption and more about pushing technological boundaries and brand visibility.
Cybersecurity and regulation: no longer optional
Cybersecurity was no longer treated as a side topic – and rightly so.
The Cyber Resilience Act will fundamentally change how embedded products are developed, maintained, and sold.
For embedded systems, this means:
- Security by design, not by patch
- Secure boot, OTA updates and vulnerability handling
- Software bills of materials (SBOMs)
- Long-term update strategies for long-lifecycle devices
Cybersecurity is becoming a core product feature and a competitive differentiator. Companies that treat it as an afterthought will struggle, both technically and commercially.
AI-supported embedded development: a quiet but powerful shift
A very important trend is not AI in embedded products, but AI for embedded development.
AI-supported tools for coding, testing, documentation and analysis are starting to impact daily engineering work. Used correctly, they can significantly improve productivity and quality. Used blindly, they introduce new risks.
The key is not automation at all costs, but augmented engineering. Vibe coding is not something that can be applied to embedded systems. This tooling will allow the experienced embedded software developer to speed up their daily work.
The intive perspective: pragmatic innovation
Across both fairs, one pattern became clear: the future of embedded systems will be driven by pragmatic innovation, not hype.
Successful embedded products now require:
- Deep system understanding across hardware, software and lifecycle
- Selective and purposeful use of AI and Edge AI. If it is not useful at all, it shall be skipped
- Cybersecurity as a foundational capability
- Tooling and processes that support long-term maintainability
As embedded systems are becoming more intelligent, more connected, and more regulated, the development process itself is undergoing a fundamental shift. Navigating this complexity now requires the strategic focus to know which ones truly matter. At intive, this is the perspective we bring to every project - ensuring that innovation always serves a clear and sustainable purpose.
.png)
.png)
.png)
.png)
.png)
.png)