Nima Kargah-Ostadi, PhD, PE, PMP
    Vice President for Research, Callentis Consulting Group
    Vice President for Project Management, Solwey Consulting
    nima@nimaresearch.com


Context-Sensitive Data-Driven Decision-Making

   
Numerous organizations have been collecting vast amounts of organizational, people- and machine-generated data at ever increasing volumes, speeds, and varieties, with various levels of quality and inter-connectivity.  My objective is to use communication and computational skills to assist data owners in extracting information out of data, solving complicated problems, and making pertinent data-driven decisions.  By pursuing a doctoral major in civil infrastructure and a doctoral minor in computational science, I have successfully developed a series of novel methodologies that have enabled me to tackle extraordinarily difficult problems.  My research approach reconciles subject matter expertise and data with computational intelligence techniques to solve complex optimization and prediction problems.  This unique tactic has led me to establish effective, efficient, and reliable analysis tools that generate sustainable, economical, and data-driven decisions regarding the transportation infrastructure networks.

In this regard, I have been using data engineering tools to acquire and prepare data.  This includes design and implementation of relational databases using structured query language (SQL), exploring data using statistical and visualization methods, and pre-processing data for scaling, feature selection, and reducing noise, variability, and dimensionality.  Following data engineering, I have employed computational science to analyze the data and report the results.  This includes solving complex problems such as classification, regression, pattern recognition, prediction, and multi-objective optimization using machine learning techniques and evolutionary computation.  During my academic and industry experience, I have developed comprehensive frameworks for benchmarking, evaluation, and comparison of various computational techniques using context-sensitive quantitative and qualitative success metrics.

In my past consulting services, I have assisted State and Municipal roadway agencies in extracting practical information out of pavement condition data to be used in pavement design and pavement management decisions.  I conducted diagnostic conversations with agencies to understand asset management needs, communicated client needs to my team, identified context-sensitive computational methodology and visualization interface to meet client objectives, developed quality assurance and quality control frameworks, and presented final products/solutions to the client and their stakeholders.

I have been applying my research paradigm to incorporate sustainability into infrastructure systems engineering and management via application and integration of computational techniques inspired by natural intelligence.  Infrastructure asset management refers to a decision support system that helps authorities in making decisions among alternatives for further development of assets or improvement of existing infrastructure.  Civil infrastructure systems are comprised of interconnected asset networks.  Increased infrastructure deterioration, increased performance demands, and challenges posed by climate change, along with limited budget and human resources have all made asset management a critical and challenging task.  The civil engineering profession is expected to act swiftly and wisely to preserve and rehabilitate existing infrastructure, and to incorporate a sustainability vision into the asset management systems.  The performance management concept in the MAP-21* and the subsequent FAST** infrastructure acts have been developed with this perspective.

Analysis tools are essential to the decision-making process at both the network and project levels.  Probably the most critical analysis tools in every asset management system are condition evaluation, performance prediction, needs assessment, prioritization and optimization toolboxes.  However, it is not easy to develop these toolboxes in a reliable fashion. Often, agencies have to deal with incomplete, subjective, ambiguous, uncertain, or erroneous information that makes the developed analysis tools, and, in turn, the decision-making process very unreliable or even irrational.  Therefore, extensive investment is required for data collection and its quality control and quality assurance precautions.  In addition, there is a need for more capable computational tools that can handle problems such as numerical intensity or ambiguous subjectivity within the collected data.

My soft computing practice employs machine learning (ML) techniques and evolutionary algorithms (EA) in addressing computational challenges of decision support systems.  ML are massively parallel computing systems capable of pattern recognition and function approximation without definition.  EAs are optimization heuristics that have good global search ability, do not depend on seed values, and do not rely on gradient information.  Computational intelligence methods, including artificial neural networks (ANN), evolutionary algorithms (EA), and fuzzy systems (FS) have been demonstrated to be promising in addressing the difficulties involved in the development of analysis tools for infrastructure asset management.  On one hand, techniques such as ML and FS have the potential to quantify sustainability indicators such as social and environmental aspects, risk and reliability, life cycle costs, and network performance.  On the other hand, EAs are powerful heuristics that can assist in multi-objective optimization of the corresponding sustainability objective functions. 


* Moving Ahead for Progress in the 21st Century
** Fixing America's Surface Transportation

Loading