Asset Modeling with Dane Boers
Welcome Dane Boers who is the founder of Modeler. Our topic will be asset modelling.
More about Dane:
Dane started his career in mining then moved to consultancy before starting Modeler. The company was started to make reliability more efficient and automated because of the repetitive work he had observed.
In this episode we covered:
- What is asset modeling?
- What is an asset class?
- Does asset modeling take into account the failure modes in the operational context; what’s impactful and what’s not?
What is asset modeling?
It is the digitization of asset knowledge. The idea is to develop a framework that combines tools such as work plans and working instructions into a structure that can be used to answer business-related questions.
The goal is to have one centralized store of knowledge about an asset class. For instance, everything about pumps are fed into one structure i.e. A pumps failure data, failure mode details, decision making logics e.tc gets captured in one place.
The information assists in answering queries related to budgeting and optimization of maintenance strategy etc.
What is an asset class?
Refers to a group of assets that share a lot of information that can be used to put them together.
Does asset modeling take into account the failure modes in the operational context; what’s impactful and what’s not?
Yes. It takes into account the failure mode and all the factors that can change the probability of that failure mode happening. Such factors might be environmental conditions and operational parameters. All these are inputs in the model.
The model then creates relationships between these factors and failure modes to provide links that can drive decisions.
So it gives an insight about what could happen and what to do about it?
Yes. It is all about statistical probability that updates predictions regarding asset failures with reasons for the same. However, there is still an element of variance and unknown.
How can companies leverage this asset model to develop effective strategy?
To understand how to leverage it requires an understanding of how the model works. It has 3 components;
- Asset model itself that contains the asset knowledge data and the logic.
- The use of the model to perform analysis or calculations to answer business queries
- Turning the answer into a digestible format i.e integrations. SAP loads, documents, dashboards e.t.c.
If it is about RCM, of importance is that all RCM related information about an asset is logically fed into the system if a query is about RCM optimization. The information could be inspections, overhauls, and subtasks to be performed. The query then could be: which of the tasks gives the best cost-benefit or life cycle cost. The model then analyses the information and gives a calculated, best alternative.
What type of data do we need?
It starts from a specific asset e.g. a pump; inputs can be the type of pump. Perhaps a centrifugal pump and the fluid that flows through the pump. Then move to the next type of pump and build a model based on the same inputs.
The inputs are then linked to turning on and off of certain components and changing failure distribution. The model improves and grows with varying operational contexts.
The basic data that you need; date of installation, equipment age, type of pump, types of fluid, impeller size etc. The model increases the need to have more data because they are connected to asset performance.
What other outputs are there apart from effective strategy?
It depends on the type of queries you have for the model. There are varying uses ranging from budgeting, equipment selection depending on operational context, intervention tradeoffs depending on the information fed in the model.
Can we update the inputs after perhaps selecting an equipment to capture the operational context?
Yes. In fact, RCA outcomes, mitigating tasks, and problem trees can be updated into the model. However, single case failures may not skew the model as much. Longer period of repetitive failures is needed for the model to produce actionable predictions.
Optimization of the model is to link back and find out what the new input means to the failure mode.
Would having a good asset hierarchy help build the model?
We tend to take a subject matter expert (SME) approach when starting out. This is because of the expert knowledge of what is important and what isn’t. However, certain important issues may not be quantified or recorded in usable data form but affect failure mode. In that case, we adjust failure rates using the SME approach.
It is preferred to start modeling at the asset level rather than system level.
Do you ever see clients experiment with different assets concurrently to see which one performs better?
Yes. Especially in equipment selection. Say; selection of electric pole materials while considering geographical, climate and environmental factors.
One thing that makes the biggest difference:
Apart from asset knowledge, linking factors to failure modes, it is essential to understand that the data in our procession is imperfect. You need to use the best assumptions in such gaps until you find better data.
The industry is moving towards data-based, prescriptive recommendations. We need to structure the data to suit this.
Dane Boers Links:
- Dane Boers Linkedin
- Modla Linkedin
- Book: RCM2 by John Moubray
- Book: The Book of Why: The New Science of Cause and Effect by Judea Pearl & Dana Mackenzie
Rooted In Reliability podcast is a proud member of Reliability.fm network. We encourage you to please rate and review this podcast on iTunes and Stitcher. It ensures the podcast stays relevant and is easy to find by like-minded professionals. It is only with your ratings and reviews that the Rooted In Reliability podcast can continue to grow. Thank you for providing the small but critical support for the Rooted In Reliability podcast!