29 January 2016

How to define BIM Use

For those that don't know, a BIM Use is a task, outcome or deliverable that a BIM model is used for. For example when a BIM model is used for structural analysis, or to create a door schedule, or provide data for an FM system.

When talking about BIM Use we mean "Building Information Models" (actual digital models), not as in "Building Information Modelling" (BIM processes). Because of this some call it Model Use, but I shall stick to BIM Use, as there is a world beyond BIM with other types of models.

BIM Use is at the core of  BIM. The basic concept of BIM is that data created is captured in a form and format that can be directly used as a resource for other purposes. 
So doors are created in such a way that a door schedule can be produced directly from those doors. That modelled structural elements behave in a way that allows for structural analysis.

Yet there seems to be massive confusion around BIM Use. What should be simple is made incredibly complicated by BIM standards, BIM contract clauses and BIM theorists.

I've written before about BIM Use and how it is being applied to LOD, in my posts LOD, are we there yet? and What is the use of BIM Use.  But those posts don't offer a solution.

Unfortunately when confronted with something unintelligible and unworkable we tend to avoid the whole thing. But BIM Use can not be ignored. If we are going to really do BIM we have to have a workable way of managing BIM Use.

The purpose of a BIM Use

Let us start with the basics. Why have a BIM Use?

A particular BIM Use must have a useful real world outcome. It should only be listed as a BIM Use for a project if there is a specific reason to do it; a specific party who will do it; a specific party who will receive the results; a specific outcome that aids the design, building or operation of the project facility.

This sounds so obvious yet is missing from most definitions of BIM Use. Discussion always seems to be around what is possible, rather than what is required, let alone practical.

Who employs BIM Uses

BIM Use is invariable talked about as the uses of external parties. Typically uses by the BIM model author are ignored.
Apparently if a structural engineer uses the architect's BIM model for structural analysis that is a BIM Use, if they use their own it is not.
I suspect this is because no cross-organisation agreement (or demand) is required it is not considered part of "BIM Process".

The problem is these in-house BIM processes are then not considered when agreeing on other BIM Uses. This can create problems when externally required BIM Uses compromise, or completely prevent, an author's own BIM Uses.

To capture who is doing what it is useful to define BIM Uses against who they are between:
  1. Within a discipline
    - e.g. schedules from model
  2. Between disciplines within a team (e.g. architect, engineer, QS etc.)
    - e.g. energy analysis
  3. Between teams (e.g. design, construction, operation, etc)
    - e.g. asset management
  4. Across disciplines
    - e.g. estimates
  5. Across teams
    - e.g. clash detection
Doing this not only ensures all BIM Uses are considered but also reveals what contractual requirements might be or not be needed for the project.

LOD is not BIM Use

For a particular BIM Use to be achievable the BIM model must have certain requirements. Currently these requirements are described via LOD descriptions. Typically an LOD has certain BIM Uses associated with it. This is the AIA [US] approach. From their E203 guide:
"The E202's Model Element Table provides a vehicle for defining Authorized Uses, Model element by Model element and milestone by milestone."
But in practice how do you define "Authorized Uses, Model element by Model element and milestone by milestone."?
Considering there are literally hundreds of different possible "Authorized Uses" are we really expected to list them not only against each Model Element, but against each Model Element at each Milestone?

The most practical LOD guide created thus far, BIM Forum's LOD specification, has tried to deal with this stipulation by kind of white-washing it. From their 2015 edition:
"Because BIM is being put to an ever increasing number of uses, the group decided that it was beyond the initial scope to address all of them.  Instead, the definitions were developed to address model element geometry, with three of the most common uses in mind – quantity take-off, 3D coordination and 3D control and planning.  The group felt that in taking this approach the interpretations would be complete enough to support other uses."  

But the AIA[US] approach is fundamentally flawed, it is the wrong way round.
BIM Uses should be listed with the required LOD against them, not LOD with allowed BIM Uses.

What LOD tables actually do is to define the level of development each element is to have as the project progresses, at each milestone.
This is a reflection of reality - project information progresses at the rate it is gathered, decided and created. You can't make information and decisions magically appear because you need it for a BIM Use, and have put it against an LOD table in a contract.

LOD specifications, matrices, tables, whatever you want to call them, need to remove references to BIM Use. It just confuses and complicates them.

A proper LOD table is an indication of model progression, when which parts will have what information available, based on what is realistically achievable.

BIM Use should be a completely separate list, referencing LOD's to describe what is required for them to be done. By comparing BIM Use requirements with LOD inclusions and progression a realistic assessment of what BIM Uses are feasible, and when they can be undertaken, is possible.

The current process of  using LOD definitions to determine what "Authorized Uses" are possible is delusional, it will never work in practice.

Who decides BIM Uses?

Another problem with the AIA [US] approach is that it defines what BIM Uses are "permitted", not what uses are necessary or even desired. Again from E203:
"The term “Authorized Uses” refers to the permitted uses of Digital Data"
Wouldn't a better approach be to define BIM Uses on a project by what uses participants want to perform? Not what a BIM author says they are permitted to perform?

In the E203 guide it states that the "usual approach" is to take the position "because some of the information is not reliable don't rely on any of it". And that their intent in E203 is to change that to "because some of the information is not reliable you can only rely on the information that I explicitly say you can."

Now that seems a sensible approach. If an architect tells you the walls in their model are LOD 200 then ignore any materials in those walls.
The problem is when LOD 200 also means the architect is saying these walls are suitable for a particular BIM Use by some other discipline. Because then we have gone from the traditional "we provide our information to you at your own risk" to "we will provide you sufficient information for you to perform your professional responsibilities."

The result of this can be the BIM author allows no Authorized BIM Uses at all, which is no better than providing it at receiver's risk.
Or the author claims information is adequate for an Authorized Use but it is not (and they refuse to rectify it), because they have no idea of what is actually required.
Or a third scenario where the BIM author is penalized (or sued) because the model they provided was demonstrably not suitable for an Authorized Use they permitted (or were forced to permit under their contract).

Either way those attempting to use the BIM model for a legitimate BIM Use are left in the lurch, and BIM authors are left at risk.

As bad as letting BIM authors decided who can do what is, there is another, worse, (and very common) approach to deciding BIM Uses. That is the assumption the owner should do it. Not only that, but the owner should do it at the very beginning of the project before the various experts required are engaged.

Of course it is legitimate that the owner make decisions on their own BIM Uses - facilities management, building control etc., and BIM Uses that may effect their decision making and built quality - crowd flow simulation, 3D visualization etc.
But asking owners to list all BIM Uses for their project is absurd. The reality is the majority of BIM Uses are by the design and construction teams, to assist them perform their work, the work the owner has engaged them to be responsible for.

Normally you would expect the owner to select design and construction professionals that have the skills to do the things they would like done. I don't understand why when it comes to BIM the expectation is that by simply listing a BIM Use in a document is will magically be done by whoever gets engaged, no matter what their skills.

I know owners are the ones that pay everyone, and so can tell everyone what to do, but that doesn't by definition make them the best qualified to make decisions about all BIM Uses on a project. Expecting them to do so is delusional.

What about Standards

For something so fundamental there is a surprising dearth of standards that directly address BIM Use. Maybe it is too much like hard work to be so specific about particular BIM Uses.

BIM Excellence.org has started a list of BIM Use definitions, 125 listed so far, although not all have actual definitions. A good start, to avoid duplication and standardize terminology.

At first sight COBie could be considered a kind of BIM Use standard. Although it sets out the required output it doesn't directly describe required model progression, and it takes no account of the LOD concept. For example it makes no distinction between data never applicable or just not available yet - any empty fields must contain "n/a" in a COBie deliverable.

A real BIM Use standard would set out what LOD requirements are for model elements to achieve the use.

The BIMforum LOD Specification is probably the only real BIM Use standard. It clearly sets out LOD requirements for quantity take-off, 3D coordination and 3D control and planning. But it should be renamed the BIMforum BIM Use Specification for: Quantity take-off; 3D coordination; 3D control and planning.

(with apologies to BIM Forum)
As BIM Use is invariably performed by software you would think software vendors would have an interest in establishing standards that optimise their software performance. Although competing software specific standards are not necessarily the best approach.
IFC is kind of in this space. MVD (Model View Definitions) define elements required for specific views of a model, which could them be used for a BIM Use. But IFC is really about software standards, not software use or BIM processes performed by humans.

I believe some standard definitions around BIM Use would be really useful. Currently beyond asking specific people on my projects I have no way of knowing what is required for a BIM Use I don't participate in.

Although standards can be part of the solution they can never be the only solution. The expectation that every BIM use for every discipline or team for every project will be covered by a standard is delusional. And what do we do while waiting for standards to be authored, discussed and agreed?
What we need are processes that establish BIM Use protocols.

Current BIM Use process

The process doesn't have to be complicated. Let's think about it from first principles:
  1. Someone wants to use something for a specific purpose.
  2. They say what that is and what they require for them to do it.
  3. Whoever is best placed to provide that is identified.
  4. Negotiations occur between the provider and user.
  5. Agreement is reached on what processes will be followed.

But in the world of BIM planning the procedure is:
  1. An authority figure decides what BIM Use everyone wants.
  2. They guess what is required to achieve these BIM Uses
    (or use a "BIM expert" to guess).
  3. They impose these requirements on everyone.
  4. BIM authors, not the owner, decide what specific information they will provide for a BIM Use.

When confronted with the obvious impracticality the usual snake oil response from BIM evangelists is that "the BIM Execution Plan is a living document that can be changed."  That might be a method to fix impractical outcomes but it doesn't justify why there is an impractical process in the first place.

A better BIM Use Process

That said negotiation is still the best method. It not only ensures everyone is doing things they are happy(ish) about, it provides an opportunity for everyone to have their say.

But negotiations have to occur in a framework that is realistic. Pretending they can occur before everyone is appointed (or that everyone be appointed at the very beginning of a project - as in IPD), or that parties will agree when there is no incentive to do so (when only authors decide what "Authorized Uses" are permissible), is delusional.

The owner should be the one to set up the framework, project participants the negotiating.

Therefore the process for owner is:
  1. The owner lists the BIM Uses they intend to do.
    - e.g. FM, budget management, etc.
  2. The owner lists possible BIM Uses that others may do, and are desirable for the project.
    - BIM Uses that may or may not be used on the project that participants may be called upon to provide BIM models capable of being utilized for. 
  3. The owner acts as arbitrator in participant negotiations.
Then as each project participant is engaged they must have shown the ability to satisfy the relevant owner's BIM Uses, and the capability to satisfy the the relevant possible BIM Uses. As the exact requirements of the possible BIM Uses are unknown, and may not even occur, fees do not need to specifically allow for them, ensuring owners are not paying for something they may never need, or that someone else (the BIM Use recipient) may pay for.

As each project participant becomes involved in the project they are required be involved in a BIM Use identification and negotiation process:
  1. BIM Use request.
    - A participant nominates what they intend to use BIM models for (including uses that the owner may have engaged them specifically to do).
  2. Define and communicate data required.
    - For each of their BIM Uses clearly describe what data they require and at what stages.
  3. Identify source/author of data.
    - Based on data required,  and through negotiation, identify who will be generating the data, or who is best placed to create the data.
  4. Agree on extent/format/form of data that will be provided.
    - Negotiate with that party on what data they can provide, and/or are willing to provide.
  5. Agree on process to supply data.
    - Negotiate timing, degree of reliance (LOD) and checking & rectification procedures.
The owner may become involved at point 3 if there is dispute over who the appropriate author is, and at point 4 if agreement on extent of data can not be reached.

If it is determined extra data is required the provider and recipient can exchange services (you scratch my back, I'll scratch yours); the recipient pays (because it saves them money); or the owner pays (it adds value to the owner and/or project). Or it is not done as there is no measurable benefit.

There you have it.
There is obviously a lot of nuance around the detail but the above process is, to me, a more realistic way of approaching BIM Use management.
It is not a radical proposal, nothing unfamiliar is introduced to BIM Execution Planning. Indeed most current BIM Planning guides would only require slight adjustment to formalize this approach.

Let's take BIM from the theorists and make it genuinely practical.