ABSTRACT
According to statistics published by the German Federal Ministry for Transport, Building and Urban Development (Bundesministerium für Verkehr, Bau und Stadtentwicklung 2006) there is a stock of about 120,000 bridges in Germany. The institutions that own these bridges are faced with the problem of maintaining these aging structures with just limited funds at their disposal (Schiessl & Mayer 2006). During the past 20 years or so, numerous so-called computeraided bridge management systems (BMS) were developed to support engineers involved in the inspection or repair planning of bridge structures (see Section 1.2): information from inspections is stored in a BMS and can be consulted at any time. This data is also used to compute the condition of the structure, described as part of a marking system. One disadvantage of many existing BMS systems is that the lack of adequate deterioration models only allows for the detection of damage through visual inspection (e.g. cracking, staining, spalling). In most cases, the best time for (preventive) repair work on concrete component parts has already passed once erosion becomes visible on the concrete surface. Predictive life-cycle management systems (LMS) are a novel approach to overcome this major drawback of conventional BMS. To detect
decay earlier on, additional non-destructive and visual inspection techniques are used in combination with fully-probabilistic deterioration models to predict the endurance and longevity of component parts. The prognosis will be constantly updated, thus being more precise concerning rate of deterioration. Such a LMS can thus be used to optimize the operation of bridges over their entire service life (Schiessl & Mayer 2006, see also Budelmann & Starck 2008). Furthermore, the system supports the institutions responsible for bridge maintenance in the long-term planning of inspections and repair measures at both bridge level and network level. Another advantage is that the required budget for funding the reconditioning measures can be planned more precisely.
