BIM and Big

 [Published in Civil Engineering Surveyor (CES) on May 2014]

We have talked about the first teetering steps, but where are the steps leading to.

BIM is frequently considered at a single project level, which is totally understandable for a discrete building, a new build for a single client. Perhaps it is the only building for that client. BIM can easily make an impact, used from the very conception of the project, through detailed design and construction and on into maintenance of the building and facilities management.

The next significant intervention will hopefully be an extension (not demolition) where the BIM model can be resurrected and dusted off, ready for reuse. There are some problems with this concept though. One is the CIC BIM Protocol seems to preclude such ongoing use. However, there are more fundamental issues. The BIM model obviously is more than just the 3D graphic image, there is all the information and data attached to it. Is that information still valid? For the initial project we gathered all sorts of new and old data together to inform the design. That data became part of the BIM model. Following construction a Health and Safety file was produced, BIM style.

Is there a process for keeping the initial information up to date, if not it becomes tired very quickly. The time is soon upon us that the data is considered unreliable and should therefore be discounted. This is a sad loss of a significant amount of knowledge. Never mind, let us re-invent the wheel again.

BIM for Infrastructure

However, BIM when applied to infrastructure has slightly different imperatives. It is still BIM and it still has rich digital data in the form of documents, digital information, and non-digital information.

The infrastructure may be in linear form such as a road, railway, or pipeline. There will be a maintenance programme in place to manage the infrastructure.

Back to the railway analogies.

What is an asset? A nut that forms part of the clamp, a fishplate, holding two lengths of rail together? A pad that the rail sits on? A bridge? A station? Or the whole network. Answers on a postcard. An asset is all of the above, form the smallest separate item through to the whole network. All of the assets have to perform together for the network, or system, to perform as planned.

Given that we have a collection of assets that form a route, part of the network, there has to be a plan as to how to manage those assets such that the route is available for use. There is, with the possible exception of roads, a direct economic construct between the maintenance of the assets forming the route and the availability of the route to earn money.

Without delving into theories on RAM, RAMS, RAMSS, or PRAMS, and asset degradation curves, and asset performance curves, for the route to be available the asset has to be maintained. Asset management becomes an imperative. (RAM; Reliability, Availability, Maintainability. RAMS, plus Safety. RAMSS; plus Security, or PRAMS; Performance.)

The route asset management plan will, of course have concepts of planned, reactive, and emergency interventions. The planned interventions may well be considered as Large Interventions, Medium Interventions, and Small Interventions. Large and Medium Interventions may well be considered as projects.

A possible exception to having an asset management plan is if you are so underfunded that the imperative is reduced to just keeping the asset running. Condition led reactive maintenance. Is this actually a prudent way to operate an asset though, is it either cost effective or safe. Let’s consider a car. Save money by not having it serviced. Saved money today, and tomorrow, only pay out when it breaks down. However, when it breaks down it is more expensive and more inconvenient. What about if that breakdown is because the break fluid has not been changed, because you skipped the service. The breaks go soft, you can’t stop in time, that’s an accident. I am not convinced that this is a viable option.

Let’s consider an enhancement driven large intervention. A Project. It is of course a BIM project. Focusing on the Information part of BIM. The project starts by gathering as much information about the existing asset as it can find. It searches for information on buried services, carries out ground investigations and topographical surveys. It commissions environmental surveys and researches the local area, SSI and heritage information. The topographical surveys may include aerial LiDAR, ground laser, traditional, precision, and physical gauges. Point clouds surveys with millions of points and millions of photographs. Some of the surveys will be repeated at different stages through the project lifecycle.

It soon becomes apparent that BIM, or PIM when only dealing with the project element is growing fast.

Also add the breakdown structures, multi-layered in a (n=9) dimensional Matrix, nD, mentioned in an earlier article.

• Asset
• Cost
• Location
• Procurement
• Work
• Interventions
• Resource
• Organisation
• And, of course, time

Then add parameters and complexities, at component level and up to project level, there is a lot of information.

All of this just at a single project level.

We also have a concept of Asset State Intervention Type (ASIT) with the following states;

• Operational
• Steady State Maintenance
• Condition improvement
• Enhancement
• New
• Redundant

How to sort this muddle

‘Here's another nice mess you've gotten me into’, to use somebody else’s catch phase. BIM is about information, but there is so much of it, in so many different ways. How can we possibly sort this muddle out? For one project, we will cope, because that is what we do. BIM will help provide some structure and discipline. Having a common data environment will require a common codification and some boxes to put all this information in.

Data collection has to be operational and proportional. It has to have a purpose. All well and good having huge amounts of paper files, but if all they do is sit in an archive box, in a nice temperature controlled warehouse, they serve little purpose. The same applies to digital information, except it is on a server, in a climate controlled environment.

So what happens at route level? We have established the need for asset management. All projects will have the same codification system. This will make them comparable. We will put all of the information from the projects into a relational database. Using BIM on projects well make this easier.

We will be able to interrogate the data using SQL. That will retrieve the information that we require. Provided we have a few people feeding the relational database, devising additional relationships, and maintaining the database, all will be well. Oh, we will need specialists to be able to create reports as well. Not open access I’m afraid.

One project becomes many. Lots of data and information quickly becomes huge amounts of data, and not unsurprising less information. The people required to handle all of that data grew to become a small army. The return on investment of that army diminished as it became more difficult to extract information from the data sets. Approaching, collecting data for the sake of data, information and knowledge lost.

If this is a probable outcome at route level after a small number of years, what of multiple routes.

The whole network is in excess of 20,000 miles, with millions of paper, negative, and cloth drawings and documents. Perhaps even the odd piece of parchment. The amount of data becomes incomprehensibly large. Every time a project is completed and the BIM data stack added to the whole information stack the system creaks. Not enough resources to manage the data. It takes so long to recover the data that it seems quicker to gather the information afresh at the beginning of the project.

Devolve the Information

The central information stack is about to fail. It is too expensive to run, does not run properly, does not give real-time information. It is no longer fit for purpose.

Solution, devolve the data and information. Take it down to route level. This will make it more manageable. Or will it. Whole data set analysis will become more difficult. Trend analysis of asset performance and establishing reliable asset degradation curves becomes much more difficult. Time to get out the spreadsheet again.

Failure looms.

There are now multiple armies of data collectors, analysts, and report writers. Lots of data but not much information getting out. Next step, cut the amount of data that we collect and retain. Discard most of it. Not exactly the rich digital sharing world we were promised with information based decision tools.

That is not the way forward.

The Asset is the backbone of our rich digital environment. Add GIS and you have your core data set.

I have just done a search in Google (other search engines are available) for TED and at the top it told me About 91,000,000 results (0.31 seconds). The top result was TED: Ideas worth spreading, which is exactly what I was looking for. (An excellent site by the way.) Those 91m sites don’t all sit in a central database. Nor do they have a common coding structure. It is all random data pulled together in real time at point of enquiry. All in less than a second.

This is how to deal with the problem of having so much digital data that it cannot be efficiently handled using traditional methods.

This is BIG DATA.

And with BIG Data come BIG Data Analytics, and Data Mining, and Data Scientists, and real information at your fingertips.

Aside -----

The volume of data is growing at a freighting rate. Google’s executive chairman Eric Schmidt brings it to a point: “From the dawn of civilization until 2003, humankind generated five exabytes of data. Now we produce five exabytes every two days…and the pace is accelerating.” Extract from BIG data; the mega trend that will impact all our lives. Also see Techonomy report in Tech Crunch.com