Artificial intelligence in bridge engineering

Artificial intelligence in bridge engineering


The idea to use artificial intelligence in bridge engineering process is very old. Since the 1950s studies on AI applications in civil or bridge engineering have proliferated. Most of these studies have dealt with specialized isolated engineering subtasks. Few of the applications have been delivered to practitioners and were used to advance their work. We analysed current situation with bridge engineering, using artificial intelligence and machine learning techniques.

Bridge engineering presents significant opportunities for AI, which exist in each of the tasks involved in the life-cycle of bridges, such as:

Decision to commission. The decision to commission or replace an old bridge draws upon many economical, political and cultural issues. These issues determine much of the overall context in which a new bridge project is situated. By understanding them one can explain, for example, which kind of bridge suits a particular area.

Design. Design of a bridge is very complex. It starts with the reconnaissance to find potential locations for the bridge together with some rough designs that need to pass some public approval.

Construction. This stage involves a separate design in which the detailed design is recreated by the construction company. Included in this design are the selection and planning of the method together with the strength calculation of the bridge.

Operation. This is basically the function for which the bridge is built.

Maintenance. This involves routine and on-demand inspection of the condition of the bridge that may be followed by the decisions to perform rehabilitation procedures, limit the bridge loads, or even close the bridge.

Replacement. The decision to replace existing bridge of a maintenance activity.

To the present day, AI systems have hardly reached a broader use in practice for several reasons. The research was often focused on narrow and specialized engineering subtasks, not on larger and more integrated problems. In addition, a building product model was not yet established and therefore not used. However, a product model, acting as a communication base, is indispensable to gain reliable design solutions, since in bridge engineering design tasks interact closely and influence each other.

Currently, a fundamental change in the Architecture, Engineering and Construction (AEC) industry by the introduction of the Building Information Modeling (BIM) technology takes place. BIM aims to represent the complete building facility in a digital product model, which is used throughout the whole life-cycle. Furthermore, parametric modeling is more and more incorporated for the design of infrastructure facilities. Apart from that, the Industry Foundation Classes (IFC), a non-proprietary data-format for the exchange of product models, has already reached a convenient level.

In bridge management, many problems, especially in bridge deterioration model, were influenced by uncertainties which not only could be solved in need of mathematics and mechanics calculations but also depend on the knowledge and experience of experts. The most interesting condition of the bridge is the performance, which is evaluated by studying the functional or structural behavior of the bridge. Bridge deterioration model is one of the key elements of bridge performance, which can be used to analyze bridge life-cycle costs and estimate the type and timing of bridge maintenance and rehabilitation needs.

Bridge deterioration model, especially deterministic models, need to update the parameter constantly when new data is collected. The conventional regressive analysis would consume a large number of computation time and computing resources. What is more, it relies heavily on the manual intervention in bridge management. However, artificial intelligence has its own superiority. It can solve this problem by means of imitate human intelligence. An artificial-intelligence-based approach (AI-based approach) for updating deterministic model parameters of bridge deterioration model was researched. AI-based approach for updating deterministic model parameters was proposed by using Bayesian theorem, which imitate some of intelligence function of human brain and realize the self-updating parameters of bridge deterioration model.

Most of the reported works on AI techniques for bridge engineering has focused mainly on design (preliminary) and maintenance because of their relative importance in bridge engineering. With help of BIM and future artificial intelligence techniques bridge construction can be pushed to another level.

Author: AI Business