Ph.D. Thesis Defense Announcement

Enhanced Construction Cost Estimation of Highway Projects using Emerging Statistical and Machine Learning Techniques

By

Mingshu Li

Advisor(s)

Dr. Baabak Ashuri (CEE/BC)

Committee Members:

Dr. Baabak Ashuri (CEE/BC), Dr. Patricia L. Mokhtarian: (CEE), Dr. Eric Marks (CEE),  Dr. Polo Chau (CSE), Dr. Minsoo Baek (CM, KSU)

Date & Time: March 29th, 3:30 pm

Location: Virtual https://teams.microsoft.com/l/meetup-join/19%3ameeting_OWNhYzA5MTUtMmJkOS00MDUzLTkwZDMtYjBmYjA2ZmYxM2U3%40thread.v2/0?context=%7b%22Tid%22%3a%22482198bb-ae7b-4b25-8b7a-6d7f32faa083%22%2c%22Oid%22%3a%2211c7e901-bf32-4624-aa0f-d3b286555d3d%22%7d

Abstract
Several state departments of transporta􀆟on (state DOTs) have encountered significant
challenges in accurately es􀆟ma􀆟ng costs for their highway projects, o􀅌en resul􀆟ng in
discrepancies between the states’ DOT es􀆟mates (owner’s es􀆟mates) and contractors’
submi􀆩ed bids. These inaccuracies can lead to cost overrun, scope change, schedule delay,
postponement, and cancella􀆟on of transporta􀆟on projects, which are problema􀆟c for both
owner organiza􀆟ons and highway contractors. There is a cri􀆟cal need to enhance the quality of
construc􀆟on cost es􀆟mates to efficiently allocate public funds and increase confidence in
engineer’s es􀆟mates. Addressing this need, the overarching objec􀆟ve of this research is to
advance construc􀆟on cost es􀆟ma􀆟on for highway projects through the applica􀆟on of emerging
sta􀆟s􀆟cal modeling and machine learning techniques, examining cost es􀆟ma􀆟on at varying
levels of granularity for a comprehensive analysis.
The study first adopts a temporal perspec􀆟ve at the monthly level, inves􀆟ga􀆟ng risk factors that
affect the accuracy of the owner’s es􀆟mate. This level of analysis allows for the examina􀆟on of
several variables represen􀆟ng the local highway construc􀆟on market, overall construc􀆟on market, macroeconomic condi􀆟ons, and the energy market to iden􀆟fy leading indicators of the
ra􀆟o of low bid to owner’s es􀆟mate. Appropriate 􀆟me-series models, such as ARIMAX, will be
applied to forecast this ra􀆟o using iden􀆟fied leading indicators. This macro-level analysis offers
founda􀆟onal insights into market trends and economic factors influencing cost es􀆟ma􀆟ons,
se􀆫ng the stage for more detailed inves􀆟ga􀆟ons.
Transi􀆟oning to the project level, the research conducts survival analysis to assess the
rela􀆟onship between several poten􀆟al drivers and the likelihood of inaccurate cost es􀆟ma􀆟on.
By innova􀆟vely applying concepts and methods from survival analysis to construc􀆟on cost
es􀆟ma􀆟on, this part of the study explores the impact of project-specific, bidder-specific, and
external market characteris􀆟cs on es􀆟ma􀆟on accuracy. This project-level analysis provides
cri􀆟cal insights into the dynamics at play within individual projects, complemen􀆟ng the broader
market perspec􀆟ve obtained from the temporal analysis.
Finally, at the most granular pay item level, forecas􀆟ng models for early-phase cost es􀆟ma􀆟on
of lump sum pay items (Traffic Control and Grading Complete) are developed using text-mining
and machine learning techniques. This approach involves retrieving project informa􀆟on
available at the early stages of project development through text analysis and examining various
machine learning algorithms with iden􀆟fied key predic􀆟ve features to select the bestperforming
model. By focusing on specific pay items, this level of analysis directly addresses the
prac􀆟cal needs of designers and cost es􀆟mators, offering precise tools for early cost es􀆟ma􀆟on
and further enriching the comprehensive understanding gained from the previous analyses.
This research contributes to the body of knowledge through: (1) developing appropriate
mul􀆟variate 􀆟me-series models (i.e., ARIMAX models) to predict the ra􀆟o of low bid to owner’s
es􀆟mate; (2) crea􀆟ng a Cox propor􀆟onal hazards model to explain and predict the likelihood of
inaccurate cost es􀆟mates; (3) developing machine learning algorithms to accurately es􀆟mate
prices of lump sum pay item at early stages of project development. It is an􀆟cipated that the
research outcome would help cost es􀆟ma􀆟ng professionals in transporta􀆟on agencies be􀆩er
understand the risk factors and poten􀆟al drivers of the devia􀆟on between owner’s es􀆟mate and
low bids, prepare more accurate cost es􀆟mates and develop appropriate risk management
strategies for enhanced decision-making. Through its mul􀆟-level analysis, the study provides
significant insights into project planning, budget alloca􀆟on, and construc􀆟on cost management,
thereby underscoring the cri􀆟cal role of integra􀆟ng machine learning and sta􀆟s􀆟cal modeling
techniques in enhancing the accuracy and reliability of cost es􀆟ma􀆟ons for highway projects.