The problem it answers
- A learning problem at the edge of the current process
- Uncertainty about whether a model is warranted
- Need for a bounded experiment
How this category works
Some of this list also lives under AI, IoT, and digital twin. We keep a single emerging view so you can see what is still early, and we cross-link instead of describing the same service twice as if it were two companies. Blockchain, spatial computing, and robotics software need a concrete job before they need a pilot.
1. Separate mature from early
A production IoT platform and a robotics experiment are not the same kind of promise.
2. Require a use
If the use case is the technology’s name, we are not ready to build.
3. Prototype honestly
A small proof with a kill criterion, then a path into a core category.
4. Fold it back
Successful work lands in AI, IoT, twin, or software — not in a permanent lab.
Questions
- Is this production-ready?
- Some items, such as applied computer vision or edge processing, can be. Others, such as a new chain or a robotics cell, start as a bounded experiment. We label which is which in the proposal.