Writing Tips - ISyE Senior Design ISyE Writing Tips...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ISyE Senior Design ISyE Writing Tips Communication matters! Communication Executives who can’t follow what you’re doing Executives won’t fund or implement the project. won’t – …and they don’t have time to try hard; it’s your and your responsibility to make it easy to understand. responsibility Sloppiness and incorrectness make you look Sloppiness BAD. BAD. – …which means you’re less likely to get raises, which bonuses, promotions, etc. bonuses, A Good Written Report Good Easy to understand – – – – Organization Explanation Readability Not “it’s all there; go figure it out” Appropriate content – What does the reader need (or not) to know? Consistent style No sloppiness – Grammar, word usage, visual details, etc. Executive Summary: Executive Mini-report, not teaser Bad executive summary (reports process, not results) The approach to analyzing and solving the problem faced by the company is multi-stepped. First, value stream mapping has been used to create a current state map that has been used to better visualize the system and the problems that currently exist. The current state map has also helped better determine exactly what data is relevant to the problem and how that data can be used to construct a solution. After careful examination and analysis of the current state, a proposed state map has been developed to show the system as it will be left. In this report, we will describe the low productivity problem in greater detail, describe the methodology behind our analysis and the way we will integrate our current work into our deliverables. TMI: Who cares?! TMI: The main data set for our project contains over 36,000 lines of data which our team imported from an Access file into SAS. The data was manipulated to read text strings into SAS and assign certain numeric values to those texts, using multiple if-then statements. For example, a new column was made labeled UNIT_TYPE and it had SAS read the first letters and gave X, Y, and Z the numeric values of 1, 2, and 3 respectively. The start time and end time of each was converted to the number of hours it occurred after the origin date of January 1, 2004. This was determined by taking the difference between the event date and the origin date and converting this value to hours. The data then had two columns dedicated to the continuous timeline from the origin date to the number of hours that have passed to the events start time and end time. TMI: What’s really important? TMI: Team Members: *George P. Burdell Joel Sokol Wayne Clough Sponsor Contact: John Smith, PhD Faculty Advisor: Dr. David Doe Location and Distance: The Sun (Approximately 93,000,000 miles / 8.3 minutes from campus) Section Type: MWF gburdell3@mail.gatech.edu jsokol@isye.gatech.edu wclough3@gatech.edu 900000000 000000000 000000000 Wordiness wastes the reader’s time Wordiness The most useful information that can be extracted from these graphs is that the demand over time is variable. As one can see from Figures 7 and 8, it is clear that the demand is variable over time. In some time periods it is higher than average and in other time periods it is lower than average. There is no way of predicting when demand will be high and when demand will be low. Although demand is variable over time, it is decreasing slightly over time for economic reasons. Figure 8 shows the variable demand with decreasing trend. The main point to take away from this analysis is that demand is variable and tends to be decreasing overall due to economic reasons. Wordiness wastes the reader’s time Wordiness The most useful information that can be extracted from these graphs is that the demand over time is variable. As one can see from Figures 7 and 8, it is clear that the demand is variable over time. In some time periods it is higher than average and in other time periods it is lower than average. There is no way of predicting when demand will be high and when demand will be low. Although demand is variable over time, it is decreasing slightly over time for economic reasons. Figure 8 shows the variable demand with decreasing trend. The main point to take away from this analysis is that demand is variable and tends to be decreasing overall due to economic reasons. Figures 7 and 8 show that demand is variable and tends to be decreasing over time for economic reasons. Now you try… Now At this point in the project we have set solid groundwork for the user interface but we will continue to build a robust system that will provide significant benefit to the facility. Through interviews with several employees throughout the facility we have been able to get a good idea of some of the major concerns of the system and we are working towards producing an interface with directly addresses (or at the very least takes into account) these needs. Now you try… Now At this point in the project we have set solid groundwork for the user interface but we will continue to build a robust system that will provide significant benefit to the facility. Through interviews with several employees throughout the facility we have been able to get a good idea of some of the major concerns of the system and we are working towards producing an interface with directly addresses (or at the very least takes into account) these needs. We will create an interface that addresses the employees’ concerns. Explanations must be understandable Explanations PCA is a multivariate statistical method that identifies patterns in data and expresses the data in a way that stresses the similarities and differences. Two main advantages of PCA are reducing the number of original variables and detecting structure in the relationships between the variables. By reducing the number of variables, PCA reduces data dimensionality which avoids the complexity of finding the patterns in data of high dimension. Detecting structure in the relationships between the variables helps transform possibly correlated variables into smaller number of uncorrelated variables called “principal components”. This transformation of data is conducted in a way such that maximum variability in the original data is reached while minimum number of “principal components” is obtained. Explanations must be understandable Explanations PCA is a multivariate statistical method that identifies patterns in data and expresses the data in a way that stresses the similarities and differences. Two main advantages of PCA are reducing the number of original variables and detecting structure in the relationships between the variables. By reducing the number of variables, PCA reduces data dimensionality which avoids the complexity of finding the patterns in data of high dimension. Detecting structure in the relationships between the variables helps transform possibly correlated variables into smaller number of uncorrelated variables called “principal components”. This transformation of data is conducted in a way such that maximum variability in the original data is reached while minimum number of “principal components” is obtained. PCA is a statistical method for identifying patterns in data that has many factors. It transforms the data into a small number of new factors (called “principal components”) that explain most of the data’s variability. Now you try it… Now The objective of this project is to build reliability models that represent the distribution of failure event types within the company’s system. Survival analysis can be used to analyze the given data. Upon receiving the data, it was further categorized into non-failure and failure events. Several graphs were created to determine the predictors of future failure events that occur. Several factors including the seasons, conditional probability, average length of failures, and the relationship between failures and events had the biggest effect on predicting future events. Our main approach will be to predict the length of a particular failure event and determine the time between failure events in the future. Progress was made in determining which factors can predict the time between failures and the length of failure events. Several functions were performed in SAS to initially analyze the survival function of generating units within the data. Now you try it… Now The objective of this project is to build reliability models that represent the distribution of failure event types within the company’s system. Survival analysis can be used to analyze the given data. Upon receiving the data, it was further categorized into non-failure and failure events. Several graphs were created to determine the predictors of future failure events that occur. Several factors including the seasons, conditional probability, average length of failures, and the relationship between failures and events had the biggest effect on predicting future events. Our main approach will be to predict the length of a particular failure event and determine the time between failure events in the future. Progress was made in determining which factors can predict the time between failures and the length of failure events. Several functions were performed in SAS to initially analyze the survival function of generating units within the data. Reliability models were used to predict the occurrence and length of equipment failure. Factors such as season and previous failure history showed the greatest predictive ability. Appendices need to be understood too! Appendices (Include explanatory text) Appendix C y = β + β x + + βn xn 0 11 β = Intercept 0 β =1 i = 1→ n −1 1 βn = O ε = Noise i.i.d Disorganized ideas are hard to follow Disorganized The company, founded in 1908, is a global supplier of refurbished and used restaurant equipment and systems. They provide restaurant equipment to companies worldwide with warehouses in the United Kingdom, and sales headquarters in the Netherlands. Their average gross sales are $125 million per year. The company’s headquarters, located in Norcross, GA, consists of an administrative building and two additional facilities. In Norcross, the company employs 180 people with 77 employees in operations. These facilities are also ISO 9001 and ISO 14001 certified. Roadmapping helps readers! Roadmapping Work Completed to Date The following section explains how data collection and research was conducted based on the three problem areas and how this information contributed to the construction of an Arena simulation. Sloppiness makes you look BAD! Sloppiness This has the potential to reduce losses and ** and drive profitability within the organization. See Appendix for a larger version. (…but there is no appendix!) Average Pr oduction Time = 153 Seconds Average Pr oduction Time/Hr = 0.0425 Hours Sloppiness makes you look BAD! Sloppiness Table of Contents 1. 2. 3. 4. 5. 6. 7. Executive Summary………………………………………… Introduction………………………………………………….. Methodology………………………………………………… Initial Results………………………………………………… Value…………………………………………………………. Next Steps…………………………………………………… Conclusion…………………………………………………… No page numbers! Sloppiness makes you look BAD! Sloppiness Table of Contents 1. 2. 3. 4. 5. 6. 7. Executive Summary…………………………………………1 Introduction…………………………………………………..2 Methodology…………………………………………………5 Initial Results…………………………………………………12 Value………………………………………………………….17 Next Steps……………………………………………………20 Conclusion…………………………………………………….21 Page numbers don’t quite line up Cutting corners looks sloppy! Cutting Page 6: The main approach that our team has chosen to design reliability models is to implement survival analysis techniques. In statistics, there are several methods in order to determine the time until a failure. One of these methods is survival analysis which is primarily used in biology as a measure to determine the time until death of organisms. The survival function generated through survival analysis can be adapted to different models such as machine failures. However, in the case of a machine failure, one has several obstacles unlike dying organisms. First, when a machine fails, this machine can be repaired and run. Additionally, one has the problem of a machine being subject to failure due to multiple causes. These are the biggest factors that impact survival analysis. Page 15: The main approach that our team has chosen to design reliability models is to implement survival analysis techniques. In statistics, there are several methods in order to determine the time until a failure. One of these methods is survival analysis which is primarily used in biology as a measure to determine the time until death of organisms. The survival function generated through survival analysis can be adapted to different models such as machine failures. However, in the case of a machine failure, one has several obstacles unlike dying organisms. First, when a machine fails, this machine can be repaired and run. Additionally, one has the problem of a machine being subject to failure due to multiple causes. These are the biggest factors that impact survival analysis. Cutting corners looks sloppy! Cutting WHAT?! Figure 14 shows that variables X1, X2, X3, and X4 are statistically significant since they have a p-value less than 0.05. Analysis of Maximum Likelihood Estimates Variable X1 X2 X3 X4 X5 X6 X7 DF 1 1 1 1 1 1 1 Parameter Estimate -0.11547 0.0005281 -0.02819 0.00351 -9.1658E-6 -0.21756 -0.01415 Standard Error 0.03172 0.0000308 0.0007749 0.0001429 0.00178 0.65063 0.05457 ChiSquare 13.2523 293.4873 1323.1594 601.7437 0.0000 0.1118 0.0672 Pr > ChiSq 0.0003 <.0001 <.0001 <.0001 0.9959 0.7381 0.7954 Hazard Ratio 0.891 1.001 0.972 1.004 1.000 0.804 0.986 Figure 14: Function Output Inconsistency looks sloppy Inconsistency He/she will monitor the results. If readings do not stabilize, he will remain at the monitor. As the company acts on its intentions to expand their operations,… The group has constructed two heuristics to find “good” solutions…. A final check is done to see if any ‘local improvements’ can be made. Sloppy grammar errors look bad Sloppy Improving productivity in the facility will amount to $26,000 and that in the receiving area will amount to $230,000. We will ensure these improvements by the standardization of process flows. The facility is experiencing growth in terms of the stores they service through December 2009. A marginal distribution for either periods of demand do not currently exist. In Europe, the Chicago facility supplies Amsterdam and Barcelona. Meaning we are taking a conservative approach and are assuming that 68% of items are expected to fall within specification. Not a sentence! Holding Cost is the Amount of Money that could be earned elsewhere if the money of that inventory if used elsewhere. Figures: what’s wrong? Figures: Figure 2.1 Good figure, good caption Good Figure 2.1 The Atlanta DC is experiencing growth in the number of stores served. The sharp increase in early 2009 can be attributed to the acquisition of Gaylan’s. “Fancy” words make you sound dumb if they’re used imprecisely One layout contains a bay utilization level of 85%. They were created to uphold all layout constraints. The heuristic determination is developed by the following operations. Productivity X has The optimality of the situation is best with 12 forklifts. X X X X XX X solution X enforce created “Fancy” words make you sound Fancy” unsophisticated if there’s a simple alternative if This insight exists from seeing the problem occur. We saw the problem occur. Analysis tools were utilized to classify and categorize the findings. We analyzed the data. The objective of the project revolves around balancing production. The objective of the project is to balance production. The purpose of this project is to increase the availability of facilities within their footprint. The purpose of this project is to increase the availability of facilities. The similarities and differences among these layouts are mentioned as well as the nature of their respective simulations The layouts and their simulations are described. Platitudes and obvious facts sound silly sound Our Georgia Tech Senior Design Team has been working to accomplish the goals of the client as well as the goals of the Georgia Tech faculty and Senior Design course. The first step to deriving any solution is to first understand the problem. Before analyzing the data, we had to collect it. Don’t make false statements Don’t Based on the average time in the system, 59.5 cars are in the system at any given time. The company’s inventory levels are increasing as sales decrease, so the company requires more-accurate long-term forecasting sales models to help them better serve their customers’ needs. It is the company’s highest priority to maximize running time during the summer. Remember chemistry? Significant digits matter… Significant WIP 0 1 2 3 4 5 CT 36.17222222 60.17222222 84.17222222 108.1722222 132.1722222 156.1722222 In 2007 and 2008, the facility handled 183,000 and 198,882 units, respectively. The number of units sold was 103,000 in 2007 and 106,452 in 2008. Total Cost Savings = $3,455.76 + $4,840.00 + $1,242.27 + $26,136.00 + $774.40 + $40,000 = $76,448.43 ...
View Full Document

This note was uploaded on 11/26/2009 for the course ISYE 4106 taught by Professor Staff during the Spring '08 term at Georgia Institute of Technology.

Ask a homework question - tutors are online