INDUSTRY & R&D
Asset degradation modeling: know how fast it wears, and why
Asset degradation modeling with TIMi estimates how the condition of an asset (battery, structure, component) evolves with use, to anticipate its remaining life instead of relying on a fixed replacement threshold, and shows the factors behind it.
Condition over time
Estimate the current state of each asset and where it is heading.
Remaining life
A trajectory per asset, not one threshold for all.
Explained trajectory
Each estimate shows what drives the degradation.
HOW IT WORKS
How asset degradation modeling works.
Three steps, in one engine, on a standard server with no cluster.
1. Prepare the data
Anatella gathers the sensor measures followed over time (load, temperature, usage cycles, depending on the asset) and the design or usage data of the asset, then structures them into variables.
2. Build the model
Modeler links the observed conditions of use to the evolution of the asset condition, from the history available on similar assets or on the same asset followed over time.
3. Estimate and explain
Each asset gets an estimate of its current condition and degradation trajectory, with the factors behind it, such as intensity of use, environmental conditions or load cycles.
Anatella
ModelerGETTING IT RIGHT
Choices that shape a degradation model.
Degradation depends on how each asset is used.
Go beyond a fixed threshold
The model makes it possible to go beyond a uniform age or cycle threshold, by using the real conditions of use of each asset. This can avoid a premature replacement, or anticipate a faster degradation than average.
Not only batteries
The principle applies to any asset whose condition evolves in a measurable way over time: batteries, mechanical structures, components subject to fatigue, depending on the sensors available.
Use comparable assets
A complete life-cycle history helps, but the model can also rely on comparable assets at different stages of their life to rebuild a representative degradation trajectory.
EXPLAINABILITY
A condition says where it is. The factors say why.
A degradation estimate alone gives a state, not what caused it. A white-box model shows the factors in the observed trajectory for this asset, so R&D and operations teams can tell a degradation due to intensive use from one due to unfavorable environmental conditions, and act on the right factor.
For R&D teams
Understanding the factors of degradation helps to guide design choices for future generations of the asset.
For operations teams
An explained trajectory lets you adjust conditions of use to extend the life of an asset instead of enduring its degradation.
For model owners
Modeler’s built-in Drift Report shows when a model loses accuracy, so you update it when needed instead of on a fixed calendar.
DATA
What data a degradation model uses.
Anatella connects to more than 80 sources, so this data can be gathered without prior custom development.
Sensor measures over time
Load, temperature, mechanical stress, depending on the asset.
Usage history
Usage cycles and load or stress cycles.
Environmental conditions
Conditions of operation.
Design and reference data
Design data or reference data of the asset.
PROOF
Degradation modeling in R&D.
Volvo is our reference for asset degradation modeling.
Isolating what degrades an EV battery
Prototypes were pushed to their limits with rapid thermal shocks. TIMi isolated the variables behind the degradation from over 1,400 sensors, feeding the design of the next generation.
FAQ
Common questions.
Is this only for batteries?
No. The principle applies to any asset whose condition evolves in a measurable way over time: batteries, mechanical structures, components subject to fatigue, depending on the sensors available.
How many sensors can be followed at once?
Volvo’s tests involved more than 1,400 simultaneous sensors. The practical limit is the capacity of the hardware, which can be sized to the project.
Does it replace a replacement threshold based on age or cycles?
It lets you go beyond a uniform threshold by taking into account the real conditions of use of each asset. This can avoid a premature replacement, or anticipate a faster degradation than average.
Do we need the full life-cycle history of an asset?
A complete history helps, but the model can also rely on comparable assets at different stages of their life cycle to rebuild a representative degradation trajectory.
RELATED USE CASES
Keep exploring.
Predictive Maintenance
Anticipate equipment failures from the data you already collect, with the reasons.
Sensor Data Modeling at Scale
Model thousands of sensor variables together on a standard server.
Equipment Failure Prediction
Estimate the probability that an equipment fails at a given horizon, and why.
SEE FOR YOURSELF
Test it on your own data.
Run TIMi on your own data with a free trial, or talk to us about your case.
