Advanced Structural Health Monitoring based on Collective Intelligence
Final Report Abstract
In this research project, advanced strategies towards decentralized, autonomous structural health monitoring (SHM) based on collective intelligence have been developed. Upon this conceptual basis, a design methodology has been proposed and validated by means of prototype implementations of SHM systems. In essence, collective intelligence has been achieved by implementing individual, collaborating software entities (“software agents”), which are (i) embedded into wireless sensor nodes deployed for monitoring and (ii) situated on remotely connected computer systems installed at distributedly located sites. In addition to the decentralized multi-agent approach, the concept of “dynamic wireless code migration” has been proposed for SHM purposes, ensuring a flexible and autonomous execution of monitoring tasks and supporting the involved human individuals pro-actively. Technically, heterogeneous hardware and software components have been developed and, through system integration, modularly been integrated into the prototype SHM systems, e.g. sensors and measuring units, database and server systems, Internet-enabled user interfaces as well as multi-agent systems, statistical analysis tools, and finite element models. As documented in several case studies conducted in this research project, the accuracy and the reliability of monitoring, as compared with existing approaches, could substantially be increased, because both local phenomena as well as global properties of the observed engineering structures are equally considered. At the same time, the resource consumption of the wireless sensor nodes deployed for monitoring could significantly be reduced. For example, a 96.4% reduction of wirelessly transferred data has been achieved, as revealed in comprehensive laboratory tests conducted on several prototype SHM systems. In addition to the laboratory tests, real-world structures have been instrumented with prototype systems to validate the newly implemented monitoring concepts in the field. Taking “clean” energy systems as an example, a wind turbine has been used as a reference structure for long-term field validation tests. Specifically, the system capabilities with respect to autonomously detecting malfunctions of sensors and measuring units (well known from real-time systems) have successfully been validated. Beyond that, it was possible to couple autonomous structural health monitoring with life-cycle management (LCM) strategies for engineering structures. As another outcome of this project, it could be demonstrated that coupling autonomous SHM and LCM does not only provide precise information about the actual structural and operational condition of a monitored engineering structure, but that it also facilitates a reliable assessment of a structure’s long-term structural performance and operational efficiency. Furthermore, owners and operators are supported in accurately scheduling maintenance and repair work at minimum associated life-cycle costs. Last but not least, the feasibility to transfer the new monitoring concepts – primarily developed for structural health monitoring – to related engineering disciplines has been investigated in this project. Preparing follow-up research in close collaboration with national and international partners, environmental engineering has been considered in more detail and monitoring solutions have been implemented, e.g., for (i) autonomous landslide monitoring, (ii) agricultural ecosystem monitoring, (iii) early warnings of slope movements, as well as (iv) soil moisture monitoring and intelligent irrigation control. As a result, valuable insights into the natural environments and their actual conditions could be achieved, serving as a reliable new basis for sustainable decision making. Nevertheless, much research is still needed to further facilitate autonomous monitoring and continuous assessment of both built and natural environments.
Publications
-
2010. A Computational Intelligence Approach towards Autonomous Monitoring of Civil Engineering Structures. International Journal of Computational Science, 4(2), pp. 113-128
Smarsly, K.
-
2010. Biologically-Inspired Condition Monitoring of Civil Engineering Structures. International Journal of Computer and Electrical Engineering, 2(4), pp. 770-777
Smarsly, K.
-
2012. A Multiagent-Based Collaborative Framework for a Self- Managing Structural Health Monitoring System. ASCE Journal of Computing in Civil Engineering, 26(1), pp. 76-89
Smarsly, K., Law, K. H. & Hartmann, D.
