During a recent weekly Twitter chat about GIS (#gistribe), a GIS industry colleague, Bond Harper, asked for some thoughts about selecting and planning an Asset Management System. Given the timing of the question, assuming GIS integration was a given. A lively discussion ensued, and we (Myself and Nathan Heazlewood) thought it worth summarizing the ideas put forth.

Overview

There are a number of ways to approach this topic, but we are going to start from a high level, and work our way down. We will go over the following:

• Requirements vs Capabilities
• Requirements
  • What is the purpose of asset tracking?
  • What assets do we need to model?
  • What information are we interested in tracking about these assets?
• Solutions
  • What are the capabilities of the asset management system you are considering?
  • Should you use a custom database solution?
  • How do you decide?
• Getting started
• Summary

As you can appreciate from these images, different types of asset may require different approaches:

Rail network and buildings	Electricity infrastructure – lines and pylons	Stormwater network

Requirements vs Capabilities

At its simplest, any technology investment decision ultimately boils down to two factors:

• What are the requirements?
• What are the capabilities of the various solutions?

That sounds easy doesn’t it?

Except it’s not.

The task of assessing the requirements for asset management is often more complicated than people imagine. Depending on an organizations needs, modeling assets properly can get very complex indeed. Compared to a lot of other GIS data/objects/features like address points or ecological areas, some types, components, and combinations of assets are much more complex. This can catch out an inexperienced GIS person when they are trying to deal with asset management.

Another portion of the asset requirements are the details about each asset. This can be complex not only because of the characteristics of the asset, but also additional related information you need to keep track of.

The flip side, of course, is the capability of the software. It is not simply a matter of looking at a checklist of features in the software and seeing if they match the asset types. Many of the asset relationships are not standard configurations out of the box. The software has to be capable of these configuration needs.
The other part of the software is pricing, and thus finding the nexus between price and capability is critical. This is where the discussion must be made of what form or type of asset management system is appropriate. Do you need dedicated software, or a more simple set of database tables and queries to serve your requirements.

Both these aspects taken together, mean this decision requires quite a bit of thought and planning. Let’s now look at each side in more detail.

Requirements

The types of asset tracking that an agency requires can be quite varied. While the location and status of an object can be tracked with an asset management system, more complex tracking is often required. Some other functional areas that are often tracked include:

• Installation history
• Warranty period
• Planned maintenance
• Unplanned incidents and faults
• Actual maintenance performed
• Valuation
• Cost of ownership
• Insurance status
• Legal/regulatory compliance
• Importance: for example is this asset a historic building (with protection from unauthorized alteration or demolition)

Many of the above aren’t simple 1:1 attribute relationships either. A good example of this is actual maintenance performed. That maintenance can take many forms, from simple cleaning, to weed control if it is an open channel, to spot repairs in various locations, or lining if it is a pipeline, all the way to complete replacement if its condition has deteriorated enough to warrant that. Keeping track of these different options requires more than a single attribute on the asset layer.

What this means is that the requirements for your asset management system are directly related to the types of assets and the information about each asset, that you wish to manage. To determine this, you need information in the following categories:

• What is the purpose behind tracking assets?
• What assets need to be tracked?
• What data is important about each asset?

What is the purpose behind tracking assets?

Before you can figure out what to track, you need to know why you are going to be tracking particular assets. The reasons are varied, and each may require specific information about the same asset. Let’s look at a few of the most common:

1. Operations and Maintenance – This is probably the key function of asset management systems, enough that they are often called another name: Computerized Maintenance Management Systems (CMMS). Whether you are a private organization or business, or a public agency, if you have physical assets, they need to have regular maintenance to ensure they are working properly. This is important for two reasons:
   1. Extends the life of said asset, reducing lifetime cost.
   2. Ensures that the asset does not pose a safety risk.
      Some items to be tracked in this area are the condition and age of the asset. You might want to keep track of maintenance history, when it was maintained, who did the work, was it regular maintenance or an unscheduled repair.
2. Condition tracking – It is worth having this as a separate entry from operations and maintenance because it can actually have applications beyond O&M. It may also be more specific than tracking for maintenance which is just looking for problems. When you are tracking condition specifically, it can be not just an ongoing record of observations, but might also be a result of a comprehensive review of an asset or set of assets. A good example of this is when a city has a street survey performed. In this process, every street is reviewed and assigned a condition score, along with an expected degradation of condition for the next 1-5 years. This can be combined with occasional observations and information about recent upgrade projects, to prioritize future work. The process to set this up may require tracking attributes for a combination of different asset types.
3. Capital Improvement Planning (CIP) – Capital Improvements is any improvement that increases the value of the asset involved. The increase in value is different from O&M where the goal is to simply maintain the condition of assets. In the case of public agencies, the planning aspect comes into play because the projects are generally grouped together to achieve economies of scale, both in time and cost. Since they are larger projects, they require more funds, and so it is important that the projects are well thought through and performed in places where they will have the most impact. As a result, there is a longer vision for planning these projects, and lists are maintained for a minimum of 5 years and sometimes extending out to 15 or 20 years, of potential projects. Keeping these lists updated and prioritized is a key place where asset tracking can come into play. A good example of this is when a city is planning a street paving project. You don’t want to just look at the streets themselves, but also any other assets in the ground, like sewer, stormwater, electrical, etc., that may need to be addressed before you pave. Nothing raises the ire of residents faster than seeing a street be paved then less than a year later, have it be dug up because a sewer or water line needs replacement.
4. Regulatory compliance – In the US, we have something called GASB 34, which requires us to maintain some valuation of each asset owned by a city, which can range from sewer to buildings to more esoteric things like trees, which are incidentally, a little recognized, but large asset, both in value and maintenance.
   There are other regulations that also require some degree of tracking assets. On the stormwater side, part of your discharge permit includes maintaining an inventory of the enclosed lines. It is just a step from there to having a cleaning schedule to ensure these lines are being maintained properly so they are clear should a storm occur.

Now that we’ve looked at some of the reasons you would need to track your assets, it is time to review the variety of assets you need to track.

What assets need to be managed?

The answer to this question may be found by looking at the items your organization is responsible for. If you are a city, this may be utility information like water, sewer or stormwater; street information, like paving, sidewalks, street lights, etc; electricity; telephone lines, poles, etc. It may include facilities, like city hall, fire departments, police stations, or larger areas, like parks and playgrounds, that are maintained by the city.
If the managing entity is a business, you might be looking at a fleet of vehicles, or real estate, both land and buildings.
If a school or university, you may find that many aspects of a city also apply, as a campus has the same infrastructure, with the addition of buildings, classrooms, and specialized networks.

The best way to figure out what is important, is to talk to the subject matter experts (SME). This may include your public works staff, facilities managers, IT staff, etc. As GIS professionals, it is easy to get caught in the trap of thinking we need to figure this out ourselves. This is folly. Ask the people that know the things. Most times, they are more than happy to tell you all about them, especially if you have something to offer that is going to help them do their jobs better. This also applies to the next topic, which is once you have figured out what assets to track, what do you need to know about them?

What data is important about each asset?

Once you have figured out why you need to do asset management, and what assets you want to manage, the next step is to figure out what is important about each of these assets. Specifically, you need to answer the question “What data is critical to maintain about these assets that will let me fulfill the purposes listed above?”

There are two parts to this process.

1. Determine how you want to represent the assets. Some assets are pretty easy to define, like sewer lines, manholes, stop signs, etc. Other assets are part of more complex networks, or have more complex components, which requires you to make some choices in representation.
   Here are a few examples:
   • Street networks: While you generally think of a street in terms of the centerline, indicating where it runs, a street also has area which is paved. Your street crew might be interested in the condition of different sections of the street, and when they were repaved. Someone working on CIP may want to be able to show the different types of projects that were done on a street, for example what was paved, or where sidewalk was put in, or different types of pavement. These would require not just a line representation, but also an area.
   • Fiber optic network: You may need to know the location of cables, cabinets, and splice cases. More advanced modeling uses may also require detail about the internal connections within those components.
     

     Wall-mount cabinet containing optical fiber interconnects. Does your asset management system need to know the details of how these connections are configured? Or, do you just need to know that there is a cable leading to a cabinet with no additional detail?

   • Buildings and other structures: You may start out with a simple building outline. If this building has sections with different owners, you may want to track that. There may arise a need to show the rooms inside a building, perhaps for emergency evacuation planning. This is additional complexity that becomes even more challenging if there are multiple floors. Some colleges and universities are starting to set up campus-wide routing so that students can more easily find their classes. This is beginning to move inside buildings as well. This means you have to also track stairs and what floors they connect. Lastly, a building may have indoor and outdoor spaces as a part of it, and these may have different maintenance needs, which is another wrinkle.
     

     One building ‘asset’? Or two? Or six?

2. Determine what attributes are necessary for these assets. This is similar to the first point except that instead of looking at the physical representation, now you need to look at the descriptive features of these assets. What do you need to capture about these assets that will provide enough detail for the purposes listed above to be accomplished? Like the assets themselves, this can range from simple to the complex:
   • Simple: dimensions, length, width, slope, height, material, date installed
   • More complex: condition, capacity of assets, most recent cip projects
   • Most complex: Ongoing condition tracking with multiple entries with timestamp; all cip projects for a particular feature; linking media to assets

With any of the above decisions, your first source of information should be the subject matter experts for each department, and also for each task area. Public works will know about utilities, and facilities, and will know the layers and attributes required to operate, maintain and improve these assets. CIP staff can tell you what is important about their projects, and what level of detail they want to have when determining assets to include in projects. Ask a lot of questions, and observe their processes, as how they perform their jobs will also help you understand what is critical to keep track of.

One common issue is when people mix too many different types of assets together, and because their asset management system is set up for one specific type then then try to fudge the design that suits that asset type onto other types of assets: for example if your asset management system is primarily focussed on say; buildings, but your organisation is also responsible for some private roads, some water tanks and some radio antennae. Sometimes users set up records talking about the ‘floor area’ of a road (but roads don’t have floors!) or the number of storeys of a radio antenna (i.e. despite the fact that antenna don’t have storeys someone has used an approximation of the number of floors to indicate the height!). Where this can cause all sorts of problems is if someone that isn’t familiar with the data starts running reports etc: issues can be created where a manager who is unfamiliar with the underlying data creates a report on the total floor area of ‘buildings’ in a region, not realising that they are inadvertently including results for many assets that aren’t buildings (but that did have a ‘floor area’ attribute).

Solutions

Now that you have figured out why you want to employ asset tracking, what assets are important, and what you need to know about those assets, it is time to find a solution for these needs. There are really two basic options: Commercial Asset Management Software, or Computerized Maintenance Management Software (CMMS); or a database level system custom to your organization. The choice depends on the complexity of what you need to track, and the amount of money you have to spend. Let’s look at each.

Commercial Asset Management Software

This is a software package that is installed on a server level, generally built on top of a relational database. It contains functionality to track a multitude of assets, and asset classes, depending on your needs. Usually these are combined into modules, which may include: sewer/stormwater, streets, parks, facilities, other utilities. The modules then contain basic attributes about each asset, including physical descriptions, length, diameter, material, etc; and also information about condition, like date installed, current condition, last maintained, etc.

Many of these systems incorporate, or are designed to integrate with a work-order system. Since the systems are tracking condition, they are often pre-loaded with regular maintenance schedules for a variety of assets and asset types. These can then be used to generate work-orders at regular intervals for maintenance. As well, if an emergency occurs, then a work-order is generated for that, and the appropriate assets are included, along with the action performed. Once this all is synchronized into the CMMS, you begin to generate a historic record of condition and action for each asset.

As with any software product, the initial look at one of these systems is going to seem like they can do everything, and include everything, possibly up to the kitchen sink. You have to make sure you take a deeper look at each system, the cost, both implementation and long-term, and the options for customization. Here are a few things to keep in mind when looking at these systems:

1. Ability to integrate with your GIS software (ideally without messy file transfers and ETL): This may seem obvious as GIS is somewhat our default, but it isn’t always. There are a lot of systems out there, and they have a range of integration with GIS, from basic, to systems that sit on top of your GIS database.
2. Ability to hold data in a way that is compatible with how GIS manages that data: One complex situation is if you currently manage linear assets using dynamic referencing. An example would be if you have sewer mains, and have sewer laterals locations calculated based on their distance along each main. Often asset management systems aren’t set up to deal with things like that. They need a permanent layer for each asset, not a dynamic layer.
3. Ability to decompose certain assets into smaller sub-assets, including knowing how many ‘levels’ of decomposition you can reach (some older systems might only let you have 4 levels of sub-assets)
4. Ability to aggregate certain classes of assets, sometimes in multiple ways: This could include combining condition estimates for different layers when doing Capital Improvement Planning, to see where multiple assets in a particular location all need work in a similar time-frame.

The other major factor to consider is cost. It is important enough to be discussed separately from the list above. If you implement a CMMS, you will have to consider a few types of costs:

1. Implementation cost – Moving your asset data from your current system, whatever that may be, and loading the new system.
2. Initial Software cost – Varies widely per system. Sometimes is based on the number and types of assets you have to manage, and sometimes is based on the size agency, or organization you are.
3. Maintenance costs – Ongoing software support. Does this include software upgrades, some level of customization support each year, etc.
4. Hardware costs – There may be requirements for a dedicated server, and the associated software needs to get that running.
5. Other software costs – Do you need/want to implement a work-order system, or a more enterprise level GIS to integrate with this system? You may need database software
6. Training/Implementation – The importance of this cannot be overstated. If you implement a software package like this, the time spent preparing your data, customizing the software to the needs of your organization, and then training staff thoroughly to use it, is critical to the program’s success and future use.
   Another key consideration is the types of ‘changes’ that may affect your asset records and how this will be reflected in your asset management system and GIS. Often in addition to the normal ‘create, read, update, delete’ operations there can often be ‘merge’, ‘combine’, ‘split’ or ‘decompose’ actions that can be difficult to deal with. For example if some of the assets that you are dealing with are something like a road centerline, and you are recording things like maintenance expenditure against road centerline records (where one line= one asset record) then what happens if it becomes necessary to split that line (because of a new intersecting road being built for example?. If you can’t tell WHERE along the original line you invested the maintenance expenditure then you can’t precisely tell how to apportion the historic expenditure against to the two (or more) ‘new’ lines.

Another issue can arise if you have used unique identifiers of a ‘parent’ asset as part of the unique identifier of any ‘sub-assets’. E.g. if there is a pumphouse building with a unique id “1234” and the ids of the pump within that pumphouse uses the encompassing building id as part of it’s it e.g. “1234-P208” then there can be a lot of problems if for some reason it is decided to move that pump to building “5678”. Do you retain the old unique id because of the historic data related to it? etc

In addition to the functional requirements of your asset management system, also consider things like the ease of integration with other systems (including your GIS), and non-functional requirements such as performance.

Custom Database System

The alternative to purchasing a software package is to develop an asset tracking solution inside your database. This option requires that you have your data stored in a Relational Database Management System (RDBMS), with spatial functionality, like Oracle Spatial, Microsoft SQL Server with SQL Server Spatial, or PostreSQL with PostGIS. This level of database is necessary because they have support for SQL views, triggers, and stored functions. The combination of these three items will allow you to perform more complex tracking of your assets.

The initial process is the same as with commercial software. You will build your GIS and structure the tables containing both spatial and attribute data as necessary to best represent the assets in the city. Once that step is complete, the next phase is to look at the types of asset tracking you want to do as described above, from maintenance management, to condition tracking, to CIP planning. You will need to figure out what special tables you would require to accomplish these.

While in some ways, this will replicate some functionality of a commercial product, the difference is that you will be able to choose exactly what is important for your organization. Here are a few things to consider:

1. Do you, or one of your staff, have the experience necessary to set up the tables, advanced queries and stored functions to create a tracking system? Setting something like this up is going to take a firm grasp of database fundamentals, as well as more advanced skills. If you don’t know it, can you learn to do it?
2. Does your GIS department have the time to spend on this sort of implementation? The system can get to be quite complex depending on your needs. Being able to think through the design of the database and then implementing that, will require some dedicated time.
3. If you have the skills and the time to put into this, is it the best use of your staff time? If the needs are complex, sure, you may be able to implement them, but at what expenditure of time, and thus delay of other projects?

As with the commercial software, cost is also a factor, but it is the cost of staff time, not the cost outlay for software. Here are the cost details to consider:

1. A custom solution implemented by internal staff moves the cost from spending to purchase software, to spending staff time dedicated to this project. How much staff time are you willing/able to dedicate to this type of development project?
2. What priorities will have to be delayed in order to implement this project? If asset management is a high enough priority, and the cost/benefit to building a system pencils out, then move forward. If it doesn’t, or there are simply too many other urgent tasks that need to be completed, you will need to consider those.
3. Generally the GIS department is funded centrally, with contributions by all the other departments, as it is meant to serve the entire organization. Given this, you need to look for buy-in from the other departments to see if this is even a priority that they are willing to have you spend time on.

How to decide?

Your decision should come down to the needs of your organization, then money and time. If you have a lot of assets to track, and are looking for maintenance and condition tracking with a work-order system, then a commercial system is going to be right. If you have just a few asset classes, and want to be able to model complex interactions between them for CIP planning, you may not be able to find that in a commercial system.
The benefit of a commercial system is that it contains a lot of asset types, has the ability to integrate with other software packages, and can maintain a lot of data about the assets in question.
The benefit of a custom solution is that it may be exactly as simple or complex as your organization needs. You won’t be paying for the overhead of modules that you never use, and can quickly add items that would not be included out of the box in a commercial package.
The last two factors, money and time, balance each other. If you have money, but no time, a commercial software package is likely the right path to take. If you have time and ability, but less available funding, then a custom solution may be appropriate.

Getting Started

Okay, you’ve looked at your organization, evaluated your needs and figured out whether to buy or build a solution. Now to get started.

The first thing to remember is to keep it simple. Don’t track more than you need to. Just because you have a lot of different data types, doesn’t mean you need to track all of them, especially at the beginning. Whether you are purchasing software, or building a solution, come up with a priority list of the assets that are most critical. Put tracking in place for these first. For a city, or agency, this will likely be your major utilities and/or streets. Sewer is often a good place to start because it is a simple system, with the potential for tracking a lot of items. For businesses, it might be your fleet. Add each vehicle, and start entering regular maintenance.

Use as much existing data as possible. If you are putting in new software, you may need to transfer data. Verify the data for quality first before you move it over. This will save error correction time in the future. For a custom solution, you will be working with your existing data, so choose the highest priority asset and build out your maintenance and condition tables, and tracking queries.

Start with the most recent data that you have. With a vehicle fleet, load in the most recent regular maintenance that has been performed. This will give a baseline that people can start working against. As time goes on, you can backfill historical data to be able to start seeing trends with maintenance and condition issues.

The last piece of advice is to not try to put everything in place at once. This applies most especially with commercial software. The company is likely to want to get their implementation contract completed as soon as possible. Have a plan in place with a timeline that breaks the asset groups into individual sections of time, as opposed to taking all the data and setting it up en masse. Doing this will enable you to go live with part of it, and test to ensure you have the attributes correct, and all the subordinate tables seeded with good data.

One thing that is really to be avoided with the above is when professionals just import into GIS and/or Asset Management Systems some poorly thought out excel spreadsheet that some business user has used for decades without putting it through a database normalisation process. A similar issue is often found if someone uses a shapefile or other flat file or de-normalised xml file and just imports that data into a GIS or asset management system without considering normalisation or optimising the structure of the data. This might make it an easy process to get started with: but it is often just storing up trouble for later.

Summary

There you have it, asset management in a nutshell. In this case, a very lengthy nutshell, but there are a lot of moving parts, and a lot to think about. Let’s run through the steps again

1. Determine the requirements of your system, starting with the purpose for tracking assets. Are you interesting in maintenance, condition, capital improvement planning, budgeting, regulatory compliance?
2. Next is to figure out what assets need to be tracked. This could be properties or vehicles if you are a business, or utilities, fleet, parks and facilities for a public agency. Start to think about how the assets and the tracking purposes intersect. That leads to the third step.
3. The details of each asset, and what data is important. This will be lead by the subject matter experts not only for each asset, but also for the tracking purposes.
4. Now that you know your needs, find a solution to meet them. Look at commercial software, or a custom database solution.
5. Commercial Asset Management software is a complete package with modules to handle many types of assets. They focus more on maintenance and value tracking for O&M and valuations. Ensure that the modules match the needs of your organization. Be aware of the costs for software, hardware, implementation, and training.
6. A Custom Database Solution is just that, custom tables, queries and functions stored in your database to handle the same functionality as the commercial software. This solution depends on your time and experience to implement, but given that, can be more flexible than commercial software. Since you determine the structure, you have the opportunity to create more tracking between disparate assets to answer specific planning questions. Time is the factor here. Do you and your department have the time to implement this solution? Are the departments you support willing to divert your time from their projects, to build this system?
7. Making the choice between systems is a matter of the needs of your organization. The complexity of your assets, the purpose for tracking them, the time you have to work with, and the amount of funding all come into play. Don’t make this choice in a vacuum. You absolutely need buy in from all departments involved.
8. Getting started is a matter of planning. Prioritize what assets are most critical. Start at the top, and implement one group at a time, testing as you go along. Bring in your most recent datasets, and maintenance/condition data to seed your tracking tables. Make a plan to backfill with older records as the historical data will help you start to see trends. Don’t try to do everything at once

As was mentioned above, taking more time at the time of implementation is going to save you frustration and tech support hassles down the road. The best advice we can give you, for all facets of an asset management implementation, is to come up with a plan and work through it. You could even use this document as a template, expanded with detail specific to your organization. This will benefit you threefold. First, you will have something to refer back to, to keep you on track. It will ensure you don’t miss a step, or forget something critical. Second, it will help you be accountable to your supervisors. You can keep them apprised of where you are in the process, any roadblocks you are running against, and items where you may need their assistance. Third, you can use this to keep your software contractor in line if you are going that route. Don’t let them jump ahead until you’ve covered everything in your process. Right now, you are probably saying, “The best laid plans of mice and men ….”, and it is true. You will need to be flexible, because things will come up as you go along, but know that you can do it!

Hopefully this has given you some insight into what we know about asset management. If you think of something we missed, see something we got wrong, want more detail, or on the off chance you think this is perfect, please leave a comment.

This article was originally published in Direction’s Magazine: Part 1, Part 2

Don’t forget to check out Nathan’s Blog!

Photo credits:

1. Image: By I, Daniel Schwen, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=3076712
2. CC BY 2.0, https://en.wikipedia.org/w/index.php?curid=2450436
3. By Канопус Киля – Own work, Public Domain, https://commons.wikimedia.org/w/index.php?curid=3628232
4. By Alby (talk) – Own work (Original text: I (Alby (talk)) created this work entirely by myself.), Public Domain, https://commons.wikimedia.org/w/index.php?curid=27148045
5. Image: By Daniel Case (Own work) [CC BY-SA 3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons