Construction industry + internet + semantic web¶

eConstruct¶

My MSc graduation project and the first 1.5 years of my PhD, I was a full-time member of the European research project eConstruct. (Sorry, no link as the old econstruct.org and bcxml.org links have died: just my own http://bcxml.net is still alive). Great experience. I learned a lot, mostly as I was thrown into the deep right away with subject experts. I was a quick learner and ended up being pretty important in the project, making one of the two main server side components, several other tools, participating and influencing core technical discussions and (co-)writing several papers (see the paper section at the end of this page). It was loads of fun and I enjoyed the company of all those other researchers.

I wrote half the contents of three of eConstruct’s official “deliverables”: official documents that explain what we did:

• bcTaxonomy. LexiCon contents and bcxml taxonomyserver documentation.

• Vitual reality information front-end (format)

• Vitual reality information front-end (implementation)

I’ve put a copy of the final eConstruct demonstration slideshow online at slideshare.net.

The eConstruct project made the decision to make the central core open source. That meant that my main contribution (the taxonomy server) and the tools I made had to be open source. I wholeheartedly agreed. So the taxonomyserver, lexicon convertor, some visualisation stuff: LGPL license. Everything is available on http://bcxml.net/. It has been some years since it was released, so it partially needs outdated versions of java servers and so. But the main idea is still there for the taking.

Graduation (MSc thesis)/ afstudeeropdracht¶

My graduation started one week before the eConstruct project had its first meeting. So from the very beginning I’ve worked in that project and also my graduation was done 100% within that project and that project’s scope. I basically did a little eConstruct of my own, so that my professor could see if any nice ideas came out of it. It did work out really well and one of the ideas now forms the basis of eConstruct, namely the use of terms defined in a taxonomy as tagnames. So if a taxonomy contains an object “door” and a property “height” the resulting file can look like this:

<door>
  <height meter="1.80"/>
</door>

Practice what you preach: I wrote my MSc thesis in xml :-) Docbook is an XML standard for books, articles, program documentation, etcetera. For those who know latex, it follows basically the same principles. Here’s the result:

• html version

• pdf version

PhD / dr.ir.¶

Weighing in at some 250 pages, here’s the end result of 5+ years of research:

• html version

• pdf version

For dynamic processes to become possible in Building-Construction, ICT support is needed. Data access and data exchange needs to be ubiquitous. Existing research often only targets elaborate international standards (that don’t really get off the ground because of international differences) or it targets big, elaborate, expensive systems (that are only available to the top 1% of the companies).

This thesis proposes an open source Building-Construction Ontology Web (bcoWeb) coupled with the so-called REST style of web services. Open source to maximise the possibility of participation and to limit the dependency on organisations or individual companies. Simple REST web interaction to keep the complexity low without sacrificing functionality. An ontology web consisting of multiple independent (national) ontologies to facilitate meaningful information exchange without first needing to build The One Big Ontology (that will never be completed).

A common theme for me is Simplicity. I’ve dedicated a separate page for some of my thoughts on that.

I defended my thesis on 15 January 2007, see the weblog entry and the photos.

MSc students I mentored¶

I was a mentor for several graduate students’ MSc thesis. Two of those thesises are for download here:

• the LexiCon by Jasper Feenstra. While at STABU he researched the diffusion and adaptation of STABU’s LexiCon as an information standard. I found it useful for my own research, as it showed that you need practical and demonstrable use of a proposed innovation before the market will jump on it. Just saying “it is for the good of the market that they’ll adopt it” isn’t good enough for them to actually adopt it.

• Instrumentatie voor het afstemmen van vraag en aanbod in de bouw (in Dutch only) by Noor Hellemans. Noor was very helpful to me in the final stage of my PhD by using my final “bcoWeb” prototype. But not only using my prototype, but actually improving the thinkwork behind it with ideas of her own. The “functional unit / technical solution”, one of the core components of my PhD thesis, was fleshed out with Noor and my professor, Frits Tolman, during many mutual brainstorming sessions.

Papers¶

Research, especially university research as a PhD, means writing papers. 8-12 pages long, typically. They force you to get your thoughts in line as you’ve got to write those very same thoughts down on paper in a coherent form. I’ve marked some of them for being good summaries of my work or my thinking.

• bcxml: the c2b and b2c scenario. 2001 CIB w78 conference in South Africa. Very early eConstruct paper. And my first conference. And a nice and impressive car ride through South Africa with my university roommate.

• eBusiness in building and construction: the eConstruct project. 2001 EBEW conference in Venice, Italy.

• bcxml: an xml vocabulary for building and construction. 2001 paper, I only contributed a bit of it.

• eConstruct: expectations, solutions and results. 2001 paper for the itcon journal.

• How bcxml handles construction semantics. 2002, CIB w78 conference (a colleague attended). Good summary of the eConstruct results with a special focus on those things that I (co-)created.

• bcxml-enabled VR project information front-ends. 2002 ecppm conference in Slovenia. Pretty good paper on web-accessible data and the REST (http+xml) web architecture. That kind of thinking was already important to me at that time and it became a core part of my PhD thesis.

• Clarity in the usage of the terms ontology, taxonomy and classification. 2003 CIB w78 conference in Auckland, New Zealand. The other side of the world. I was getting fed up with the different ways in which some people I worked with used those terms. So I figured out the common usage by going through all 200 papers of a previous conference. For me, the definitions I cooked up still stand.

• Semantic web technologies applied to building specifications. 2004 CIB conference in Toronto, Canada. Good summary of the final stage of my PhD. Using the web and RDF to link construction information. Focused on building specifications, as they have many inherent information linking needs.

• Practical use of the semantic web: lessons learned and opportunities found. 2004 ecppm conference in Istanbul, Turkey. Less practical than the above Toronto paper; focused more on developing my theory on ontologies.

Presentations are generally not as useful as a paper, but I have uploaded most of them to slideshare.

I’ve made good summaries (yes, I’ve been told by several others that they’re good!) of several conferences and other events. You can find them on my weblog tagged as “AEC”.