W. M. Kim Roddis

RetroLite: An artificial intelligence tool for lighting energy-efficiency upgrade

A prototype software tool integrating knowledge-base and database approaches is developed to solve lighting-retrofit problems for energy conservation and management purposes. Founded on the types of problems and the conceptual differences between conventional programming and knowledge-based expert systems, it is felt that a system which combines both approaches is best for solving lighting-retrofit problems. The architecture of the system is built on the concept of knowledge-based expert systems and its links to a database. A commercially available knowledge-engineering tool and a dBase III-compatible complier are used for system development. Improvements necessary to upgrade the current research system into a deliverable tool are suggested.

FATIGUE CRACK INVESTIGATION FOR THE ARKANSAS RIVER BRIDGE IN HUTCHINSON, KANSAS

Abstract Fatigue cracks were identified in the Arkansas River Bridge at: (1) stringer to floor-beam connections; (2) expansion joints; and (3) floor-beam to exterior girder connections. Cracks at floor-beam to exterior girder connections are caused by out-of-plane distortion, and are the major focus of this investigation. A review of the crack phenomena identified in the bridge, evaluation of fatigue stresses at the crack details under traffic, analyses for two possible repair approaches to prevent further crack propagation, and recommended retrofit methods to extend the bridge life are presented. Both the observed crack features and the proposed repair methods are evaluated by a finite element analysis. The maximum stress range after the repairs decreases by 50% to a level where cracks would not be expected to reinitiate.

UNCERTAINTIES OF ASPHALT LAYER THICKNESS DETERMINATION IN FLEXIBLE PAVEMENTS-INFLUENCE DIAGRAM APPROACH

Abstract The thickness of pavement layers is an important parameter used in Pavement Management Systems (PMS). Thickness data are used for pavement condition assessment, performance predictions, selection of maintenance strategies and rehabilitation treatments, basic quality assessment, and as input to overlay thickness design. Pavement thickness is usually determined from direct testing such core samples, nondestructive testing such as radar, or historical records such as pavements network database. This paper proposes the use of Bayesian Influence Diagrams as a tool in providing a probabilistic model for thickness determination procedure in flexible pavements. The Bayesian Influence Diagram Model is presented as a framework for addressing uncertainties involved in capturing quantitative and qualitative information in the asphalt layer thickness determination procedures. The model is also used to perform value of information analysis in the determination of pavement layer thickness. The Influence Diagram...

Knowledge-based assistants in collaborative engineering

Collaborative engineering requires the exchange and use of design information in a multidisciplinary team with time and space separations among the designers over the course of the design process. A challenge to effective collaboration is the delivery of existing knowledge to the time and place required. Upstream communication is necessary when participants later in the sequence, such as the construction team, possess information needed by those earlier in the sequence, such as the design team. For separate entities sequentially performing design and construction, knowledge-based assistants can provide such essential information as relative costs of construction choices when the construction partner is not yet identified. For a single entity team performing design-build, knowledge-based assistants can provide consistent and thorough implementation of design choices considering cross-disciplinary and construction issues. Steel building design/fabrication/ construction serves as an example to provide specific context for discussion of the use of knowledge-based assistants with emphasis on upstream communication of design information. Cost savings are used to measure the value of such knowledge-based assistants. Knowledge-based assistants encapsulating cross-disciplinary knowledge can also be used to assist in design situations encountered on a semi-regular basis by a design organization but infrequently by an individual designer. Cases and rules in combination can give designers guidance based on pooled organization experience. Such use of knowledge-based assistants is briefly illustrated for 1) steel bridge fabrication error resolution and 2) repair of fatigue damage found for in-service bridges.

Overcoming the Language Barrier: AISC's CIS/2 Initiative

A common language for electronic data interchange is a key enabling technology that has been largely lacking in the construction industry. The inability to systematically share project data among the members of the constructed facilities team has been a major barrier to delivering advanced computing and information technologies. Adopting a standard object-oriented format for the electronic exchange of data enables the building and sharing of product models. The American Institute of Steel Construction’s recent endorsement of the CIMSteel Integration Standards (CIS) has lead to pioneering projects using object-oriented product models in the design, detailing, procurement, fabrication, and erection of steel framed buildings. Actual benefits in terms of time and quality are examined. Schedule reductions are quantified, showing substantial time savings of about one-quarter, even in a pioneering environment. By examining the evolving use of CIS/2 in the North American steel building industry, a grounded perspective is gained from which to consider the successes, failures, barriers, and solutions for transfer of information technology advances from research to application in practice. The product data model as key enabling technology The construction industry is infamous for its degree of fragmentation both vertically and horizontally. This industry fragmentation has lead to an information technology (IT) infrastructure classically characterized by “islands of automation”. That is to say, commercial engineering software products address a part of the constructed facilities product or process, but there is no provision for systematic interaction and integration of the isolated individual implementations. This lack of a “common language” has proved a persistent barrier to realizing in practice the possible benefits of more advanced computing approaches. The need for standard product and/or process models has been well known for many years. Despite many research demonstrations of feasibility and several major standardization efforts, progress has been markedly slow. A recent bright spot in the arena of application of an open data exchange standard to the construction industry is the rapidly growing use of CIMSteel Integration Standards, Version 2 (CIS/2) in the North American steel building industry. This provides an opportunity to quantify the benefits realized by the practical deployment of this critical enabling technology. Recent pioneering projects

One-Way Reinforced Concrete Flexural Elements: Economic Characterization and Selection Guidelines Developed Using Adaptive Simulated Annealing

Adaptive Simulated Annealing is used to generate economic cost maps for one-way slabs and beams subject to flexure and shear which are used to establish revised guidelines for economic selection of concrete flexural elements. Using #6 to #11 reinforcing bars produces unnecessarily expen sive designs. #3 to #7 bars were present in the overwhelming majority of the optimal designs. For girders, two layers of flexural reinforcing appeared more frequently in optimal designs than did a single layer. The expression bd 2 = 20 Mu. does not provide a generally accurate means of proportioning the least expensive flexural members. The least expensive members frequently had reinforcement ratios significantly lower than the assumed 1%. For girders, stirrups were very often included in optimal desig ns. For one-way slabs, small (#3, #4, #5) and medium (#6, #7, #8) bars are used more frequently in optimal designs. When deflection is not calculated, medium bars appear most frequently in optimal slab designs.

Repair of Fatigue Cracking due to Out-of-Plane Distortion

Out-of-plane distortion-induced fatigue cracking is caused by relative rotation and displacement between longitudinal girders and transverse members framing into these girders. This paper describes distortion-induced cracking observed in Kansas Department of Transportation bridges, reviews the current design details used for connecting transverse and longitudinal members, summarizes different approaches that have been used during former repair practice, and presents case studies of three representative bridges that are under current investigation.

Machine learning in updating predictive models of planning and scheduling transportation projects

A method combining machine learning and regression analysis to automatically and intelligently update predictive models used in the Kansas Department of Transportation’s (KDOT’s) internal management system is presented. The predictive models used by KDOT consist of planning factors (mathematical functions) and base quantities (constants). The duration of a functional unit (defined as a subactivity) is determined by the product of a planning factor and its base quantity. The availability of a large data base on projects executed over the past decade provided the opportunity to develop an automated process updating predictive models based on extracting information from historical data through machine learning. To perform the entire task of updating the predictive models, the learning process consists of three stages. The first stage derives the numerical relationship between the duration of a functional unit and the project attributes recorded in the data base. The second stage finds the functional units with similar behavior—that is, identifies functional units that can be described by the same shared planning factor scaled in terms of their own base quantities. The third stage generates new planning factors and base quantities. A system called PFactor built on the basis of the three-stage learning process shows good performance in updating KDOT’s predictive models.

Using Artificial Intelligence for Concurrent Design in the Steel Building Industry

The ability to communicate clearly is vital to any quality engineered project. Because of the high degree of interdependency among activities in the concurrent design of complex engineering systems, the ability to coordinate and synchronize communication among design team members is crucial. The computer-based tool, SteelTeam, is under development and integrates knowledge-based expert systems with a computer-aided design and drafting environment to create a system that helps bridge communication gaps in the steel building industry. SteelTeam makes design, fabrication, and construction knowledge available to each of the parties involved in the steel building design. In addition, SteelTeam provides a standard format for the design information that is passed between parties during the design process. The completeness of this information is crucial to the generation of design alternatives, which are necessary to ensure that the optimum product is developed. By looking at the issues involved in concurrent engineering, it is possible to identify areas whose improvement leads to faster product development, lower cost, and improved productivity. These issues are addressed by SteelTeam through improved communication between the parties involved in the design of the steel building.