Once an architectural-engineering-construction (AEC) project gets launched, it keeps moving like a rocket through space. Only in extreme cases (e.g., when money runs out) does a project come to a screeching halt. Due to tight budgets, interest payments, re-mobilization costs, etc., construction projects can’t afford to just stop when obstacles are encountered. Instead, alternative materials, methods, designs, or any other resource use must be considered to keep the project on track.
Projects often get off track. This is due to a variety of reasons, many of which relate to 'tolerances.' In this proposal, the term tolerance is used broadly to denote any variation from the point solution that was designed and planned. Tolerances are the natural consequence of the manifestation of uncertainties. They are a fact of life. The impact of tolerances may be dealt with by requiring craft workers to take the heat or by means of ad-hoc, remedial planning, but these solutions are costly as well as time consuming. More proactively, tolerances may be controlled by selecting more sophisticated machining or assembly methods. Alternatively, the product and execution processes may be designed more robustly so that tolerances will impact them less.
The need addressed by this proposal boils down to this: AEC practitioners as well as academics lack the vocabulary and conceptualization to describe tolerances, yet tolerances significantly affect performance in this industry. This lack is reflected in our practices, teachings, and software tools, including those for 4D CAD modeling, which are inadequate in terms of reflecting alternative futures. To alleviate this problem, the proposed research will investigate how performance risk due to tolerances can be reduced through the application of new principles combined with software solutions. Some principles being considered have already been applied in manufacturing. They involve the application of distributed set-based design for improving design fitness throughout the project life cycle and analysis of the allocation of tolerances based on production method capabilities.
The proposed research will investigate various kinds of tolerances encountered during AEC product/process development and how they affect system performance. For each tolerance, the key research question is "Which project stakeholder(s) is (are) in the best position to manage that tolerance?" The objective of the research is to develop a system for tolerance allocation. A method is needed to decide where and when in the product/process development effort the tolerance has an impact, and correspondingly, to decide how this situation is best managed.
In contrast to contractual approaches that are commonly used in the AEC industry to allocate risk, the proposed research will investigate means for making designs and schedules more robust, so they will withstand impact better. The PI will study representation and reasoning schemes that make it possible to model potential alternative futures. As opposed to deciding on point-based solutions at each step in the product development process based on local optimization criteria, the research will develop set-based solutions and carry them forward up to the last responsible moment, when more global system optimization criteria can be considered.
In this proposal, the term design fitness is used to denote the ability of a design and a schedule (product and process) to withstand the impact of tolerances and the corresponding changes while adhering to the project goals. Design in this context is a broad term that refers to the development of the product itself (i.e., design in the narrow sense) but also the development of processes for production (manufacturing, fabrication, and construction), operations, maintenance, and decommissioning.
Expected results include (1) identification of tolerances in AEC products and sensitivity of processes to tolerances, (2) case studies with simulations illustrating the impact of tolerances and tolerance allocation on project execution, (3) a methodology for modeling tolerances and for conducting tolerance allocation in the AEC product development process, (4) review of existing software tools to describe their expressiveness with respect to modeling tolerances, (5) if appropriate, development of guidelines for new software capabilities (6) a methodology for capturing product and process decisions that allow for alternative futures, (7) a decision support system to trade off multiple objectives between project stakeholders in order to gauge design fitness.
Findings on Tolerance Management in the AEC Industry
Please refer to the website on TOLERANCES.
Publications
| Prof. Iris D. Tommelein | E&PM Program | U.C. Berkeley |
E&PM Program / IDT / 6 September 2004