1. ATU Decision Support System

2. Overview Decision Support System – what is it? Definition Main components Illustrative Scenario Ontology / Knowledge Base Development Purpose of the ontology Main content of the ontology Progress Future plans

3. Decision Support System (DSS) DSS: a computer-based information system that supports business or organizational decision-making activities. ATU DSS: an interactive software tool that supports asset management decisions by integrating and reasoning with diverse information sources about different assets and their relationships.

4. DSS – What it is NOT! It is not a tool to tell expert asset managers how to do their job! Asset managers are experts in their fields and have a variety of existing systems to help with asset management. What they are not experts on is the effects of their actions on other assets, soil, roads etc.

5. E.g.: Decisions on Buried Assets Change operational parameters Add new buried assets Replacement Emergency repairs Abandon Wait and monitor Diagnostics (more surveys) Combine with other works (road, buried assets) … An asset manager will know what they WANT TO DO in the context of their own assets This may not necessarily be the best course of action in terms of the global asset ecosystem

6. Potential Users of the DSS However – The purpose of our RESEARCH is to demonstrate integrated management of an asset ecosystem (Road, buried assets, soil, etc.) is beneficial.

7. Main Components of the ATU DSS Knowledge Base: ontology and datasets Cost Models Deterioration Models User Interface

8. Knowledge Base of the ATU DSS Ontology: defining main concepts of buried assets, ground conditions, the environment, human activities and their relationships. +Real world datasets: existing authoritative mapping/geotechnical/utility databases Ordnance Survey of Great Britain British Geological Survey UK Water Industry Research Sensor data continuous monitor sensors Sensor based surveys (including ATU sensors)

9. Deterioration Models of the ATU DSS Deterioration models can be used to predict the remaining service life of utility asset or time of failure assess the asset condition (condition rating) or the risk of an asset failure predict failure/break/deterioration/vulnerability rates prioritise replacement and rehabilitation.

10. Cost Models Compare the outcomes of different options for decision makers Suggest a cost efficient option. Evaluating the outcomes regarding Social impacts Environmental impacts Economic impacts

11. Illustrative Scenario: Leaking Water Pipe A water utility company notices a minor leak in a pipe. This is not an urgent case (not a burst) and a decision has to be taken whether to dig a hole to fix the pipe. The utility company has appropriate data about the assets it deals with but does not have data about other related assets that may affect the condition of the water pipe or can be affected by the pipe deterioration. The ATU DSS aims to provide asset managers with an integrated view of the local asset ecosystem.

12. Illustrative Scenario: Leaking Water Pipe What knowledge and data must be in the DSS to support such a scenario?

13. Ontology + Reasoning An ontology defines a common vocabulary for people who need to share information in a domain. It includes machine-interpretable definitions of important concepts in a domain and relations among them. An ontology can be used to perform multi-step reasoning to evaluate outcomes of actions and suggest courses of action. Leaky water pipe  [Causes] Soil changes  [Results in] Loss of support for road surface and nearby assets  [Results in] Reduced lifetime of road surface and nearby assets  [Can be modelled by] Method X  [Which requires] data Y and Z

14. ATU DSS Ontology The ATU DSS ontology defines the main concepts describing buried assets (e.g. pipes) soil land cover (e.g. pavement) the environment and human activity Methods for deterioration modelling, cost modelling and sensing/data acquisition. as well as their relationships.

15. Why this approach? Isn’t this just common sense? Can’t we just do this with guidelines? Why do we need a computer system? A: Complexity! …

16. A diagram representing (part of, simplified) relationships between soil and pipes This is only ONE SMALL PART of the entire knowledge base

17. Main Content of the Ontology Knowledge & Knowledge Acquisition techniques

18. Implemented Ontology Components Reusing SWEET ProcessPhenomena Human Activity SubstanceProperty SWEET generic concept ontology developed by NASA, defining general earth and environment terms https://sweet.jpl.nasa.govhttps://sweet.jpl.nasa.gov

19. The Property Ontology

20. The Phenomena Ontology

21. The Process Ontology

22. Implemented Ontology Components - based on standards OntologyStandard Land CoverNational Land Use Database: Land Use and Land Cover Classification (Version 4.4, 2006) Land UseNational Land Use Database: Land Use and Land Cover Classification (Version 4.4, 2006) SoilGeotechnical investigation and testing — Identification and classification of soil (BS EN ISO 14688)

23. The Land Cover Ontology

24. Pipe Crack Causes and Influences

25. Ongoing Work and Future Plans Developing a sub-ontology describing properties and processes of soils. – will help answer questions regarding how soil is influenced by the environment and how soil influences water pipes. To finish development of the ATU DSS ontology; To implement a prototype of the ATU DSS; To evaluate the ontology and the prototype DSS system in the “to dig or not to dig” scenario. Thank you!

27. The following slides are spare.

28. Decisions on Roads Wait and Monitor Resurface Reconstruct Patch repairs Overlay …

29. Illustrative Scenario: Leaking Water Pipe

30. To Dig or Not to Dig?

31. Specialization for Water Pipe and Road

32. Reasoning with the ATU DSS Ontology Statements within the Ontology: –  means `possibly affects’ and it is transitive – WaterPipeLeak  SoilWaterContent – SoilWaterContent  SoilStrength – SoilStrength  SoilTrafficability Question: Does WaterPipeLeak possibly affect SoilTrafficability? Answer and Explanation (generated by a reasoner) – Yes – Using the statements above, we have: WaterPipeLeak  SoilTrafficability