- SAP Managed Tags:
In this blog article, I want to take a closer look on two major concepts of asset management. First, I want to explain how to share asset information across the asset life cycle and how to collaborate and enable new business models with the help of an asset collaboration platform. Second, I want to give an overview of maintenance strategies that makes the Asset Management more cost efficient.
What is an asset collaboration network?
An asset collaboration network, like SAP Asset Intelligence Network, provide easy mechanisms to share data beyond companies borders. They are not only for sharing static asset data but also constantly changing dynamic data, like for example sensor data. The advantage of the network is, that is not necessary to create a new data set, instead you pass along the data across all stages of an asset life cycle. The following example illustrates how they can be used:
A manufacturer sells a physical asset to an owner and with it the data of the asset. This can be a price, the weight but also data like 3D models, error codes or digital manuals. The owner of the asset can now instruct a service provider to install the asset. Therefor the provider can use the information which got shared by the asset owner. Additionally, it can add more important information to the owner’s data pool, for example a location where the asset got installed. The asset owner can use the data from the manufacturer and the service provider to implement for example an asset strategy. At any time, participants of the asset collaboration network can contribute to a single data source of an asset owner. Besides new ways of collaboration, the network allows also new business models like selling or renting asset data.
Technically the asset collaboration network is implemented by cloud database. Data adapters help to bring the data in a uniformed structure, so they can be shared. This allows the manufacturer and the service provider to collaborate with the asset owner even if they use different asset management systems. The asset owner decides how to handle the shared data. The first option is to decline the sharing request of a third party. The other option is that the owner takes on the data. Then he can decide whether the shared party have still the option of changing the data or whether the shared party should have access to the data in the future. An asset collaboration network is a powerful tool to build one reliable asset data set, where also third parties can contribute.
How to define an Asset Maintenance strategy?
Maintaining assets can be very cost intensive at the same time can a failing asset have tremendous impact on a business. Companies who can lower their maintenance costs while improving the asset reliability have a decisive advantage over competitors. Maintenance strategies can help organizations to optimize resources, lowering lifetime costs and enhance reachability of an asset.
There are various types of asset strategies. The graphic shows an overview on how asset strategies can be classified. Corrective maintenance (CM) is the oldest and simplest way of maintenance. A service happens only when an asset fails. Time based maintenance describes a strategy where maintenance activities get carried out in temporal periods. Condition based maintenance does maintenance if a diagnostic procedure concludes that there is a need for action. The most modern strategy is the reliability centered maintenance (RCM). It combines all strategies and helps defining the right maintenance activities for an asset. Therefor RCM looks at two parameters, the condition, and the importance of an asset. The condition can be defined by rating criteria like error rates, age of an asset, or on base of sensor data. The importance can be evaluated by impact on the environment, company image lost, or cost arise by an asset failure. After the rating of importance and condition an RCM-matrix helps to find the right maintenance strategy. The matrix suggests for an asset with low importance and good condition CM, for an asset with good condition and high importance TBM and for an asset in a bad condition with low importance CBM. When an asset has bad condition and is highly important RCM suggests additional analysis like Reliability Centered Maintenance Analysis (RCMA), Failure Modes and Effects Analysis (FMEA) or Root Cause Analysis (RCA), to name a view. This analysis looks at different functionalities of assets, how these could disturb, what impact such a failure mode could have, and how it could be avoided or predicted.
What is predictive Maintenance?
Prediction plays a major role in modern asset management. The so-called predictive maintenance is a method to support RCM and CBM. Models need to be used to be able to predict when a maintenance activity is necessary. To build up model, indicators must be defined. Information about the indicators can be gathered by sensors. Machine learning algorithms can now find patterns in the data to predict the asset failure. Two ways of machine learning training methods are common in asset management. The supervised training links labels of a normal and bad conditions with a data segment of the data set. The algorithm is then able to predict misbehavior of an asset. The second training method is called unsupervised learning. This investigates how an indicator influence the overall system. The algorithm can now predict the failure of an asset when certain indicators misbehave.
A good maintenance strategy and the possibility to predict asset failures are a powerful tool to optimize the asset management of an organization. In the next blog entry, I want to take a closer look on how to use mobile devices and asset visualization tools to make maintenance activities easier.
What is an asset collaboration network?
An asset collaboration network, like SAP Asset Intelligence Network, provide easy mechanisms to share data beyond companies borders. They are not only for sharing static asset data but also constantly changing dynamic data, like for example sensor data. The advantage of the network is, that is not necessary to create a new data set, instead you pass along the data across all stages of an asset life cycle. The following example illustrates how they can be used:
A manufacturer sells a physical asset to an owner and with it the data of the asset. This can be a price, the weight but also data like 3D models, error codes or digital manuals. The owner of the asset can now instruct a service provider to install the asset. Therefor the provider can use the information which got shared by the asset owner. Additionally, it can add more important information to the owner’s data pool, for example a location where the asset got installed. The asset owner can use the data from the manufacturer and the service provider to implement for example an asset strategy. At any time, participants of the asset collaboration network can contribute to a single data source of an asset owner. Besides new ways of collaboration, the network allows also new business models like selling or renting asset data.
Technically the asset collaboration network is implemented by cloud database. Data adapters help to bring the data in a uniformed structure, so they can be shared. This allows the manufacturer and the service provider to collaborate with the asset owner even if they use different asset management systems. The asset owner decides how to handle the shared data. The first option is to decline the sharing request of a third party. The other option is that the owner takes on the data. Then he can decide whether the shared party have still the option of changing the data or whether the shared party should have access to the data in the future. An asset collaboration network is a powerful tool to build one reliable asset data set, where also third parties can contribute.
How to define an Asset Maintenance strategy?
Maintaining assets can be very cost intensive at the same time can a failing asset have tremendous impact on a business. Companies who can lower their maintenance costs while improving the asset reliability have a decisive advantage over competitors. Maintenance strategies can help organizations to optimize resources, lowering lifetime costs and enhance reachability of an asset.
There are various types of asset strategies. The graphic shows an overview on how asset strategies can be classified. Corrective maintenance (CM) is the oldest and simplest way of maintenance. A service happens only when an asset fails. Time based maintenance describes a strategy where maintenance activities get carried out in temporal periods. Condition based maintenance does maintenance if a diagnostic procedure concludes that there is a need for action. The most modern strategy is the reliability centered maintenance (RCM). It combines all strategies and helps defining the right maintenance activities for an asset. Therefor RCM looks at two parameters, the condition, and the importance of an asset. The condition can be defined by rating criteria like error rates, age of an asset, or on base of sensor data. The importance can be evaluated by impact on the environment, company image lost, or cost arise by an asset failure. After the rating of importance and condition an RCM-matrix helps to find the right maintenance strategy. The matrix suggests for an asset with low importance and good condition CM, for an asset with good condition and high importance TBM and for an asset in a bad condition with low importance CBM. When an asset has bad condition and is highly important RCM suggests additional analysis like Reliability Centered Maintenance Analysis (RCMA), Failure Modes and Effects Analysis (FMEA) or Root Cause Analysis (RCA), to name a view. This analysis looks at different functionalities of assets, how these could disturb, what impact such a failure mode could have, and how it could be avoided or predicted.
What is predictive Maintenance?
Prediction plays a major role in modern asset management. The so-called predictive maintenance is a method to support RCM and CBM. Models need to be used to be able to predict when a maintenance activity is necessary. To build up model, indicators must be defined. Information about the indicators can be gathered by sensors. Machine learning algorithms can now find patterns in the data to predict the asset failure. Two ways of machine learning training methods are common in asset management. The supervised training links labels of a normal and bad conditions with a data segment of the data set. The algorithm is then able to predict misbehavior of an asset. The second training method is called unsupervised learning. This investigates how an indicator influence the overall system. The algorithm can now predict the failure of an asset when certain indicators misbehave.
A good maintenance strategy and the possibility to predict asset failures are a powerful tool to optimize the asset management of an organization. In the next blog entry, I want to take a closer look on how to use mobile devices and asset visualization tools to make maintenance activities easier.
