Faculty of Engineering, Transport & Logistics ManagementTheses / Dissertations submitted to Department of Transport & Logistics Managementhttp://dl.lib.uom.lk/handle/123/127182024-03-28T17:48:09Z2024-03-28T17:48:09ZAn Optimization model for multi-objective vehicle routing problem for perishable goods distributionFernando Mhttp://dl.lib.uom.lk/handle/123/216672023-11-17T21:32:40Z2022-01-01T00:00:00ZAn Optimization model for multi-objective vehicle routing problem for perishable goods distribution
Fernando M
Vehicle Routing Problem (VRP) is a well-studied area of operations research that has resulted
in significant cost savings in global transportation. The primary goal of the VRP is to find the
best route plan that minimizes the total distance traveled. The current study used VRP to solve
the problem of fresh Agri products distribution in retail chains. With the advancement of
computation power, researchers pay more attention to incorporating real-world characteristics
when developing VRP, making it more practical for use in real-world applications. Existing
literature identifies a research gap in richer problems that use real-world characteristics
concurrently. This study created an integrated bi-objective VRP model that focused on
resource optimization, order scheduling, and route optimization all at the same time. Two
objectives aim to minimize distribution costs while ensuring product deliveries to retail outlets
on time. To improve real-world applicability, the model incorporated multiple real-world
characteristics simultaneously. All the algorithms were developed using an open-source
optimization library called OR-tools.
This research compared several heuristics and metaheuristic methods respectively, to obtain
the IBFS (Initial Basic Feasible Solutions) and iterative improvements. Thereafter, best
performing heuristic method (savings algorithms) and metaheuristic method (guided local
search) were hybridized to develop the proposed two-phase solution method. All the solution
algorithms and the developed VRP model were tested using the data obtained from one of the
largest retail chains in Sri Lanka. Numerical experiments show the efficiency of the proposed
solution algorithm in solving a real-world VRP problem. Further, numerical experiments show
that the proposed VRP model has achieved a 16% saving in daily distribution cost while
ensuring on-time deliveries to 95% of the retail outlets. Further, on-time deliveries of fresh
Agri products ensure the freshness conditions. The developed VRP model is efficient to use
as an operational planning tool for planning distribution operations in retail chains.
2022-01-01T00:00:00ZRetail sales forecasting in the presence of promotions : comparison of statistical and machine learning forecasting methodsChamara HHHRhttp://dl.lib.uom.lk/handle/123/216662023-11-17T21:32:41Z2022-01-01T00:00:00ZRetail sales forecasting in the presence of promotions : comparison of statistical and machine learning forecasting methods
Chamara HHHR
Retail sales forecasting is the process of estimating the number of future sales for a specific
product or products. However, producing reliable and accurate sales forecasts at a product
level is a very challenging task in the retail context. Many factors can influence observed sales
data at the product level, such as sales promotions, weather, holidays, and special events, all
of which causes demand irregularities. Sales promotions are one of the salient drivers in
generating irregular sales patterns. Sales promotions confound retail operations, causing
sudden demand changes not just during the promotion period, but also throughout the demand
series. As a result, three types of periods are relevant for sales promotions: normal,
promotional, and a post-promotional. However, previous research has mostly focused on
promotional and normal (i.e., non-promotional) periods, often neglecting the post-promotional
period. To address this gap, we explore the performance of comprehensive methods, namely
gradient-boosted regression trees, random forests, and deep learning in all periods. Moreover,
we compare proposed approaches with conventional forecasting approaches in a retail setting.
Our results demonstrate that machine learning methods can deal with demand fluctuations
generated by retail promotions while enhancing forecast performance throughout all time
periods. The base-lift model outperformed machine learning methods, although with more
effort necessary to cleanse sales data. Our findings indicate that machine learning methods can
automate the forecasting process and provide significant performance even with the standard
approach. Hence, our research demonstrates the way retailers can successfully apply machine
learning methods in forecasting sales.
2022-01-01T00:00:00ZOptimization applications in seaport container terminal operations and fuzzy logic-based inter terminal truckingWeerasinghe BCAhttp://dl.lib.uom.lk/handle/123/222032024-02-07T21:59:23Z2022-01-01T00:00:00ZOptimization applications in seaport container terminal operations and fuzzy logic-based inter terminal trucking
Weerasinghe BCA
Operations research techniques have helped optimize container terminal
operations over the past decades and have been a regular feature of maritime logistics
and maritime supply chain literature in addition to being in practice at container
terminals across the globe. The first phase of the project, a systematic review
systematically collates through Scopus and analyzes 1631 papers published in the
domain to find the main research clusters and understand future research directions.
Studies based on both quayside and landside planning are encapsulated for this
research. Five research clusters that discuss simulation, scheduling and automation,
quayside operations, integrated operations, and container transportation are identified
based on author keywords of the systematically derived paper pool. In addition to that,
the evolution of applying optimization techniques in container terminal operations
planning is discussed in this study alongside the suggested trajectory of the research
agenda under each cluster. The analysis finds that genetic algorithms, integer linear
programming and heuristics are the most widely used operations research techniques
in container terminal optimization. The review proposes the application of methods
such as neural network, fuzzy logic and deep learning models related to artificial
intelligence to widen our understanding of container terminal operations. The second
phase of the project is conducted to optimize ITT truck flows in container terminals
by following the research direction that is derived through the cluster, container
transportation in the systematic review. The proposed model is developed using fuzzy
logic in MATLAB software. The model can allocate ITT trucks based on the demand
from terminal yards and current truck arrival rates at gatehouses. More industry and
academia as well as inter-terminal collaborations are needed in future studies for
enhancing ITT operations and the overall operations in container terminals.
2022-01-01T00:00:00ZTransport infrastructure project planning process and capital deployment efficacy implications : evidence from Sri LankaAbeysekara GMBPhttp://dl.lib.uom.lk/handle/123/214082023-10-13T01:43:48Z2021-01-01T00:00:00ZTransport infrastructure project planning process and capital deployment efficacy implications : evidence from Sri Lanka
Abeysekara GMBP
Public sector investments contribute to economic growth and enhance the productivity of a
country. According to the Central Bank annual report 2020, the Government of Sri Lanka
spends around 30% of total annual expenditure on capital investments. The project planning
process plays a crucial role in selecting projects and allocating resources with optimum capital
efficiency. However, public investment planning in Sri Lanka does not consider returns on
investment or technical specifications, resulting in low value for the cost. It also lacks expert
consultations or in-depth evaluations of project need. As a result, implementors often fail to
adhere to the process of selecting capital-efficient ventures.
It is important to comprehensively evaluate projects at the planning stage and select the most
cost-efficient options. This analysis of the current project planning process in Sri Lanka is
undertaken based on selected case studies of road and rail transport infrastructure investments.
Sri Lanka allocates a share of its public capital investment for land transport infrastructure
projects. This research aims to examine the planning process for public sector capital
investment projects, identify its implications for capital efficiency and propose efficiency
improvements to Sri Lanka's public sector capital investments.
Standard operating practices in Australia, Canada, Germany, Hong Kong, India, and Vietnam
were compared with Sri Lanka. This comparison revealed that Sri Lanka's public project
planning process does not provide standard operating practices for initial screening of projects,
clientele analysis, probabilistic risk analysis, or post-project evaluation.
Context analysis from rail and road infrastructure investment projects revealed that most
projects in Sri Lanka do not follow the current project planning process. Instead, project
proponents bypass the process and seek direct Cabinet approval. On average, projects that
receive such direct approval have higher investment costs associated with them. In-depth
interviews with industry experts revealed that the causes of capital deployment inefficiencies
were the unclear institutional role entrusted by the Department of National Planning, political
influence, lack of transparency throughout the process, and gaps in the operation manual.
Upon summarising all gaps and identified issues, the study proposes appropriate mitigation
measures. These measures recommend developing an Act of Parliament to introduce an
authoritative body to implement the planning process and adopt centralised guideline with
Standard Operating Practices. The proposed centralised guideline was validated through two
rounds of expert consultations. Future research could evaluate the proposed centralised
guideline by applying them to specific projects and expanding their application to encompass
irrigation, port, and airport infrastructure investments.
2021-01-01T00:00:00Z