3. 3
Typical industrial sensor in 2020
Vibration sensor (up to 1,000 times per second)
Temperature sensor
Water & explosion proof
Can send data >10km using 25 mW power
Processor capable of running >20 million
instructions per second
4. 4
But... what does it actually do?
Once an hour:
• Average motion (RMS)
• Peak motion
• Current temperature
5. 5
99% of sensor data is discarded due to
cost, bandwidth or power constraints.
https://www.mckinsey.com/~/media/McKinsey/Business%20Functions/McKinsey%20Digital/Our%20Insights/
The%20Internet%20of%20Things%20The%20value%20of%20digitizing%20the%20physical%20world/The-
Internet-of-things-Mapping-the-value-beyond-the-hype.ashx
8. 8
On-device intelligence is the only solution
Vibra&on pa+ern
heard that lead to fault
state in a weekTemperature
varies in a way that
I've never seen
before
Machine
oscillates different
than all other
machines in the
factory
12. 12
TinyML
Inspired by "OK Google"
Focus on inferencing, not training
Machine learning model is just a mathematical
function with lots of parameters
Accuracy vs. speed, reducing parameters, hardware-
optimized paths
Targeting battery-powered microcontrollers
Pete Warden
Neil Tan
13. https://www.flickr.com/photos/oceanyamaha/7091324605
13
What is it good for?
Recognizing sounds Detecting abnormal vibration
https://pixabay.com/photos/washing-machine-wash-cat-4120449/
Biosignal analysis
https://www.flickr.com/photos/sheishine/16696564563
Anything with messy, high-resolu3on sensor data
14. 14
Enabling new use cases
Sensor fusion
http://www.gierad.com/projects/supersensor/
20. Classification
What's happening right now?
Anomaly detection
Is this behavior out of the ordinary?
Forecasting
What will happen in the future?
20
3. Letting the computers figure it out
21. 21
Picking the right algorithm
Classification
Neural network
Anomaly detection
K-means clustering
Forecasting
Regression
23. 23
Edge Impulse
Embedded or edge
compute deployment
options
Test
Edge Device Impulse
Dataset
Acquire valuable
training data securely
Test impulse with
real-time device
data flows
Enrich data and
generate ML process
Real sensors in real time
Open source SDK
24. 24
Fitting in a wider ecosystem
Leveraging TensorFlow Lite for microcontrollers (neural networks)
CMSIS-NN for optimized machine learning operations
CMSIS-DSP for optimized DSP operations
... with the ability to use additional resources from silicon partners
CMSIS-NN