Projects

Box-Plant Trucking Logistics

Full Description

Task #1 Preparation for Exploratory Case Studies

Because logistics operations in the containerboard mills and box plants can be complicated, distributed, and not well documented, focusing on specific products, such as containerboard or corrugated boxes, and select two or three base operations to conduct targeted research. After a series of discussions with Dr. Liker and Dr. Jacquelyn McNutt of the CPBIS we determined that this project should concentrate primarily on the trucking logistics around corrugated box plants. This decision enables us to use similar analysis techniques in facilitating future comparisons of the logistics practices in the paper and auto industries. For example, the unit of observation used in the auto study may be applicable in our study.

The unit of observation divides a larger study component, such as a warehouse, into small elements that allow for more specific analysis. In the case of the auto industry, a unit of observation is a particular product that is moving from a supplier to a plant. That is, if a supplier sends the same product to two different plants, this constitutes two units. If a supplier sends two different products to the same plant, that is also two units. This allows for a potentially more in-depth analysis and larger statistical sample. This definition could also be true in the paper industry.

For our project, selection criteria for operations to be included cover:

1. The likelihood of getting permission to interview operations' personnel and collect truck transport logistics data;
2. The size and structure of logistics activities that characterizes the geographic distribution of operations and their customers;
3. Whether the operation is part of a company that is integrated - this will facilitate the study of the impact of firm partnership in their logistics performance;
4. The ease of collecting needed information;
5. Whether a distribution center or cross-docking is used; and
6. Whether advanced logistics planning tools are used in making trucking load and route decisions.

Ideally, among all operations' candidates with a similar trucking transportation logistics environment, we would like to include at least one operation that uses its own trucking fleet and another that uses third party logistics providers. This will provide the team the opportunity to understand the pros and cons in outsourcing the trucking logistics activities. Our goal is to understand their logistics' environments, activities, personnel, decision processes, and costs -- in order to explore opportunities for improving the operations' logistics efficiency and cost.

As a first step, general insights on these issues will be obtained from two sources: literature that currently exists and in-depth discussions with Jacquelyn McNutt, the Executive Director of the CPBIS. Thus, our research can start before the approval of the human subjects policy issues, which sometimes can take a while to go through the approval processes.

Drs. Lu and McCarthy will work with Dr. Liker's team to assist us in implementing our research strategy. For example, the survey of American auto parts supplier trucking logistics performance designed by Liker's team (2002) will be used as a starting point in designing a survey for the paper industry. When our team visits operations' sites, we should be clear what records are needed, what types of questions should be raised, if some of the detailed proprietary information cannot be released, can their aggregates or covariate information be available to us, etc for retrieving some of the needed missing information.

Task #2 Science-and Observation-based Profiling Studies of Logistic Records

We propose the following "hierarchical" procedure for the profiling research of logistics data. For example, consider a specific time period of trucking operations. At the highest level, the profile study will aggregate the weight of materials transported or mileage serviced by all trucks. At a more detailed level, one can profile only the trucks from a particular operation, e.g., a firm's specific mill to a box plant (for particular types of products). Then, at an even more detailed level, one can profile against an operator providing the service. The summary information (e.g., weights or mileage) at successive time periods can be studied through time series modeling techniques. Profiled data at different departments or sectors of the integrated company can be compared for efficiency improvement (e.g., lead-time) or cost reduction reasons.

Good analysis of logistics data requires successful data collection practice. In order to limit efforts in tracking down details of logistics information at distributed locations, we propose to use the following "reverse-engineering" idea to design data gathering schemes. First, using the information obtained from Task #1 to decide case study goals and needed data and their analyses to support the study goals.

For example, in studying the efficiency of trucking utilization (one of the performance evaluation criteria in Liker's survey (2002)) at the firm level, we need aggregates of weight of materials transported through certain units of mileage and lead-time (and possibly other normalizing factors). Then, with proper permission we will track down trucking logistics records for collecting data meeting our goals. Data analyses and report findings will follow. This approach will be more efficient than the conventional approach of getting data first without knowing what to do with it, especially in the case that the data size could be huge and "incomplete". The risk is that the study goals might not be completely meaningful after analyzing some of the collected data even with the experience learned from the survey in auto industry. We expect that there will be several iterations between goal setting, data collection and analysis, and conclusion drawing (in the hypothesis-testing format).

Task #3 Profiling Box Plant's Decision and Logistics Operation Environments

Suppose that from the profiling research, we can identify ways of improving an operation's trucking logistics operating performance. However, the operation might be reluctant to accept the suggestions of using certain analysis to make logistics decisions. For example, box plant decision-makers potentially do not have the proper tools (e.g., optimization algorithms or data analysis procedures) to find better strategies for their logistics operations. Often, different departments of an operation do not communicate with each other and thus their strategies are developed "locally," not optimizing "global" (cross-departmental) goals. Thus, conducting profiling research requires understanding an operation's "culture" and its logistics environment, and the processes of planning logistics. This is particularly important in the investigation of the impact of using various alliance structures in logistics operations.

In this task, due to its scope, we do not necessarily expect a complete quantitative analysis of operations' decision profiles. Several qualitative measures, such as are the trucks used in one department made available for use in other departments when they are not fully utilized, or do the customers (e.g., corrugated box plant managers) view the "quality" of logistics operations as one of the important performance indicators, could be used to aid the quantitative measures collected in Tasks 1 and 2 for this profiling study.

Task #4 Successful Models for "Best-in-Class" Logistics

With the information collected in the previous three tasks, we are confident that our team can identify traits and characteristics leading to successful (and unsuccessful) trucking logistics operations. With rich enough data, a "model" for a "best logistics practice" could be constructed. In this project, this model represents a collection of excellent activities to build a successful trucking logistics operation. A quantitative regression type of model might not exist at this stage. Through comparisons with the "benchmark" from the best practice, an operation can trace down the weakness in the environment (or decision process) of its logistics for a more focused study to find improvement tactics.

Benchmarking successful trucking logistics cases in other industries could also be beneficial. There are several conference presentations, articles and books of surveys on distribution and logistics operations in various industries such as apparel industry (Rajamanickam et al., 1998), food, and beverage industry (Cottrill, 2001), etc. Our team will retrieve information comparable to paper industry's operations for cross-industry studies.

Task #5 - Design an Industry-wide Survey

Learning from the experience of the on-going survey conducted in the auto industry, we will design an industry-wide survey to gain a better understanding of the paper industry's trucking logistics activities. This will allow us to study the cost structure, efficiency of truck utilization, customer service level, and quality of logistics operations industry wide. A comparison of their performance with other industry's trucking logistics activities could provide proper benchmarks for recommendation of potential improvement.

The following summarize potential publications from this one-year project.
1. A case study report/paper to describe the current practice of an operation's trucking logistics operations from case studies.
2. A report/paper presenting the success factors in the corrugated box plants' trucking logistics practices.
3. A report studying the impact of using different alliance structures in mills and corrugated box plants for their trucking logistics operations.

Furthermore, there is a great chance for organizing a special issue in Transportation Research, Part E. Papers from our research of the trucking logistics issues in the paper industry (explicitly the corrugated box segment) and Dr. Liker's studies on the similar issues in the auto-industry are excellent candidates for a special issue publication.

We will work with both the CPBIS and TIP Centers' industry sponsors and the Centers' management to ensure that the research project teams activities are consistent with all existing CPBIS/TIP protocols governing the collection of data and the dissemination of research results through reports, papers, and conference presentations based upon the data gathered from this project. We expect to present key project ideas in an appropriate conference for sharing project information and promoting scientifically based resource planning for logistic activities.

Overall, the case studies will help operations selected to better understand strengths and weaknesses of their trucking logistics operations. Comparisons among operations with trucks owned by the firm and third-party logistics providers could provide pros and cons associated with outsourcing a firm's trucking logistics operations. These studies together can help operations develop strategies for improving their trucking logistics efficiency and quality and for reducing cost and delivery time.

Specifically, in working with company mentors and the CPBIS/TIP Centers' management teams closely, our project team will produce various types of reports. The collected data, information summary and research results are useful to company personnel and to the Centers' administrators for planning possible follow-up research. More importantly, the process of working out all details of the stated project tasks is an educational instrument for actively involving industry people at many levels (engineer, finance officer, managers and directors, etc.) in reviewing their decision processes of planning logistic activities. This will pave the way for possible future implementation of suggested tactics and organizational structure changes.

The data collection and analysis required for this project will be primarily accomplished by students involved in the graduate programs for the CPBIS and TIP and under the supervision of the Principal Investigators. The students will be trained in technology competence, economic modeling and information synthesis (including statistical analysis) skills, and system engineering knowledge (process flow modeling and logistics planning) to conduct project investigations. The students graduating from our group will make significant contributions to the pulp and paper and trucking industries based on their training, interactions with company representatives in this project, and understanding of the trucking logistics operations.

Many of the data analysis tools will be of high value to the trucking and pulp and paper industries for understanding their processes better. For example, the profiling methods and data analysis tools and examples can be used as tools for the industries to analyze their operations' efficiency and identify areas for improvement.

Our data and tools can also be useful to other groups in conducting their research. For example, researchers focusing on enterprise effectiveness could use our data in their follow-up studies of optimizing a box plant's logistics activities.

Duration: 1 year