TinyML
Definition
TinyML runs machine learning on small embedded devices with low power and memory.
Machine learning on highly resource-constrained embedded devices, typically microcontrollers with tight power and memory budgets.
Simply explained
Even a tiny computer in a machine can run a learning model, for example to detect unusual noises directly on site.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A machine builder detects bearing noises locally on a microcontroller without permanent cloud connection.
Limits and pitfalls
- Small models are not automatically robust or secure; memory, power, sensor aging, and update path must be measured.
Origin & history
TinyML emerged from embedded ML and gained a common comparison field through standardized tiny benchmarks from 2021.
Use in mid-sized companies
- Possible application: A machine builder detects bearing noises locally on a microcontroller without permanent cloud connection.
Frequently asked questions
What does TinyML mean?
Machine learning on highly resource-constrained embedded devices, typically microcontrollers with tight power and memory budgets. What should you watch out for? Small models are not automatically robust or secure; memory, power, sensor aging, and update path must be measured.
Sources
- MLPerf Tiny Benchmark · Banbury et al., MLPerf Tiny Benchmark
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: