Laiserin's LemmaLessons from the Plant Kingdom
(lemma: a short theorem used in proving a larger theorem)
Not long ago, I was invited to speak about "Lessons Learned in Collaborative Design & Construction" at a gathering of 2,000 petrochemical and pharmaceutical process manufacturing customers of Aspen Technology at their annual AspenWorld conference. My hosts expected to be enlightened by what they imagined were the myriad successes and advanced art of AEC project collaboration. Instead, I learned yet again that it is the AEC industry taking the hindmost, while other industries go quietly and methodically about building knowledge.
From the opening plenary on "Engineering and Process Lifecycle Management" to topical sessions such as "Streamlining Workflows for Collaborative Engineering" and "Virtual Plant Simulators," AspenWorld presentations focused on leading edge technology implementations for plant and process industries that most folks in AEC don't dare dream of. Even more impressive was the high proportion of presentations delivered by end users reporting on their own experiencesa sort of "been there, done that" which far outstrips the usual AEC presentation mode of "like to go there, wish to do that."
The session I contributed to was called "Internet Experience in Engineering Collaboration and Asset Management," and included panelists from Jacobs Engineering, Dow Chemical, Intergraph Corporation and Statoil.
For starters, Jacobs reported 1-3% communications savings, 2-5% savings on rework, and 10-20% reduction in total engineering costs derived from the internally developed Jacobs Project Interface (JPI). Rejecting ASP providers as an inappropriate model for mission-critical infrastructure, the self-hosted JPI project collaboration system is built on Lotus Notes database technology and Domino servers with Cimmetry Jvue technology for document viewing and redlining. Of course, with nearly 40,000 employees and over US$4-billion in annual revenue, Jacobs enjoys technology development advantages beyond the reach of most AEC businesses.
Dow Chemical reported on a pilot project using ASP project hosting pioneer e-Builder. Lessons learned included: lower travel cost (via 3-D model review); cycle-time reduction (through web-accelerated document approvals); direct reduction of repro and mailing costs; and a 20% savings in project admin time to manage documents and transmittals for issue and review. All impacts were measured through a rigorous "six sigma" methodology.
Intergraph's SmartPlant system embodies the kind of integrated model-centric collaborative design process that is generally lacking in commercially available solutions intended for the AEC market. SmartPlant uses a "publish and subscribe" system organization to integrate data from multiple (third-party) applications and multiple models. The result is a shared project model with change control in the hands of design engineers (not programmers or system administrators) across multiple organizations.
Statoil, the Norwegian oil company, developed a change management system for design and construction of its North Sea drilling platforms. Creating each of these multi-billion dollar structures can involve 120,000 tagged items, 60,000 documents and 8,000 changes. Statoil found that 80% of its in-house users needed only to view (rather than create or edit), with a strong preference for views generated from data rather than from drawings. By using Oracle's CADview3D technology, Statoil easily achieved two-way integration between 2-D schematics and/or 3-D models on one hand and tags and/or data on the other (the technology underlying CADview3D was originally launched in the mid-to-late 1990s as EnVision by Adaptive Media, which became Vuent, which was acquired by i-Engineer, with the product being renamed Envision-i, which then became Assentive Solutions before being sold to Oracle).
In contrast to the hard evidence and innovative technology of these plant/process examples, all I could point to on behalf of AEC collaboration was: the reassuring but non-rigorous user satisfaction survey by and for Constructware's eponymous ASP service; the slightly more rigorous but still limited ROI study conducted by the now-defunct Hurwitz Group on behalf of Autodesk's Buzzsaw hosted system; and the otherwise estimable Cyon Research white paper on a collection of transportation infrastructure projects using the Citadon ASP servicea paper unfortunately mooted by Citadon's subsequent adoption of an entirely new-from-the-ground-up system for collaboration, document management, viewing and so on.
If there's a take-away here (other than the obvious, "the grass is always greener..."), it's that AEC design and delivery processes are unlikely to improve until AEC users and their technology providers insist on and actually perform rigorous metrics on statistically valid samples of project experience. To quote the classic process/change management cliche: what isn't measured can't be changed. If that bar seems too high (or too low), let me know what you think.
Editor and Publisher, The LaiserinLetter
Analysis, Strategy and Opinion for Technology Leaders in Design Business