what are the benefits? critical path non-critical activities
what are the inputs? tasks sequence/dependencies durations
Sequence the most common sequences / dependencies Task ATask B Task A Task B Task C Task B Task A Task B depends upon Task A; B cannot start until A is finished Task C depends upon Task A and B; C cannot start until both A and B are finished Tasks B and C depend on Task A; neither can start until A is finished, but B and C are independent of each other
rules and conventions just one start just one finish no danglers task number and/or name duration early start time late start time early finish time late finish time float slack
the process - 2 stages draw analyse Task 01Task 02Task 04Task 06Task 03Task 05Task 07Task 08Task 09
more unusual links and relationships so far all links have been finish-start links... Task ATask B Task A Task C Task A Task B depends upon Task A, but with a 3 day delay; B cannot start until 3 days after A is finished The finish of Task C depends upon the finish of Task A The start of Task C depends on the start of Task A; this is a start-to-start link; it may also incorporate a delay 3 days
using the outputs Gantt Charts resource histograms resource smoothing optimising the schedule
Scheduling task 1 task 2 task 3 task 4 task 5 task 6 task 7 task res.Duration Calendar 2 days 3 days 1 day 2 days 7 days 4 days 1 day Jane Bill Jane Jim Bill Jane Bill
PERT and uncertainty - 1 Task 01Task 02Task 03Task 04 4 2 5 4 The critical path looks like tasks 01-02-04 But what if you are not sure about the duration for Task 03?
PERT and uncertainty - 2 Task 01Task 02Task 03Task 04 4 2 5 4 The better estimate for Task 03 might be its PERT estimate, or Expected Value, EV) EV = O + 4L + P ___________ 6 O = Optimistic estimate (say 2) L = Most Likely (say 4) P = Pessimistic (say 12) In this example the EV = 6, which does, in fact change the critical path
PERT and uncertainty - 3 But how confident can we be in these results? An durations spread is the degree to which estimates of the duration differ from each other. If every estimate of duration were about equal, the estimate would have very little spread. There are many measures of spread. The distributions on this page have the same mean but differ in spread: the distribution on the bottom is more spread out.
PERT and uncertainty - 4 Standard deviation is used as a measure of spread. In a normal distribution, about 68% of estimates are within one standard deviation of the mean and about 95% of the estimates are within two standards deviations of the mean. EV = O + 4L + P ___________ 6 O = Optimistic estimate (say 2) L = Most Likely (say 4) P = Pessimistic (say 12) SD = (P – O)/6In our example SD = (12-2)/6 = 1.666 So, we could say that, for task 03: With 68% certainty, the duration will be between 2.34 and 5.66 (4 ± SD) With 95% certainty the duration will be between 0.68 and 7.32 (4 ± 2SD)