Assignment

Answers to the questions should be in your own words; do not copy or reproduce material in constructing your answers. Be specific and precise in your answers.  Use a bulleted sentence(s) format where possible.

1. St. Joseph Health (SJH) is a system of hospitals in California and other western states.  Recently, their patient throughput rate has significantly declined causing cash flow problems. 

A computer simulation of patient flow pointed to the time that it takes to "turn" a patient room after a discharge is a significant factor in the throughput rate. 

SJH administrators feel that it is currently taking too long to clean and restore a room once a patient vacated the room, and have set a target of 55 minutes for this cleaning.  In addition, they defined a "nonconformance or defect" as any turnover that takes longer that 70 minutes or less than 30 minutes (based on concerns for proper cleaning).

A study has been completed that captured times for an Environmental Services team to clean and restore rooms in 10 samples of three random hospitals each (drawn fromall SJH facilities). See the table below.

Since this is the only such data they have, they do not know the overall population standard deviation and therefore must rely on conversion tables in setting up charts.

Question 1.a:  Based on this limited sample, Discuss the quality of their process for room cleaning; is it in a "desirable state" from an Consistency standpointAssume that SJH is currently a "3 sigma organization" in their quality rigor and standards.  In your answer,

1) First, what is the appropriate control chart(s) to use here?  Why; defend your choice.

2) Develop the appropriate charts.

a. First, provide a chart(s) thatignore a chart format that includes recognition of patterns (i.e., provide the "basic control chart(s)" discussed early in our lecture notes on SPC Control Charts)

b. Provide a second version of the chart(s) that does allows for patterns to be identified; see the latter slides in our SPC Control Charts lecture notes for this format.  

3) Base on the 10 samples available and the charts you develop, what conclusions can you draw from these charts?

4) How sure (%) are you of your conclusion?  What is the Type 1 error of your conclusion?

5) Is it safe to provide management with a prediction of the time to clean rooms in the near future?  If so, what would that time be?  If not, why not?

6) ASSUME that your prediction/estimate is the target, 55 minutes a room.  If in a typical 8 hour shift there are 140 rooms to be cleaned, how many cleaning crews would be needed for a shift?

2. Due to customer complaints concerning errors in deliveries and long delivery times, Sulliavan Calculator Inc has decided to enhance the current circuit board component system used in a certain calculator to add some redundant components at key places. 

The current system has 5 components in series with the following reliabilities: 0.90, 0.95, 0.98, 0.90, and 0.99.  The new system would provide a backup to the first and 4^th components.  Each of these would have a liability of 90%.

a. What is the reliability of the current calculator design?

b. How much improvement does adding the backups give?

3. Shell Oil conducted a designed experiment aimed at increasing the speed with which oil wells can be dug.  Eight factors were considered:  (A) flow rate, (B) rotational speed, (C) type of drilling mud used, (D) drill load, (E) type of coolant used, and two others as shown below.  Two levels for each factor were selected as shown in Table 1.  Level 1 is the current method used by Shell.

Table 1

      Factors                                                                        Levels

1                                                                    2   

(A) Flow Rate                                     low                            high

(B) Rotational Speed                            low                            high

(C) Type of Drilling Mud                        cast iron                     steel

(D) Drill Load                                      low                            high

(E) Coolant Type                                cheap                         expensive

(F) Bit Design                                     current                        new

(H) Mud Pressure                                550 psi                         650 psi

An initial Taguchi type experiment was set up using the simple 7 Factor, 8 Trials matrix (see lecture slides).  The results were questioned by several experts, so Shell decided to conduct a second experiment to confirm the original conclusions. The result of each trial/run in the second experimentwas as follows (the performance measure is drilling rate measured in centimeters per second).

Questions:

1. Based on the "Column Effects Method" of analysis, discuss the effects of the various  factors on their ability to impact the rate of drilling for the new results.

1. How would you recommend that this drilling process be conducted to optimize the drilling process (in terms of settings of each factor)?   Why?