Project Management Resource Scheduling Eng. Giorgio Locatelli

Transcription

1 Resource scheduling Project Management Resource Scheduling ng. Giorgio Locatelli Mauro Mancini Mauro Mancini Resource Scheduling Project Management: The planning, monitoring and control of all aspects of the project and the motivation of all those involved in it to achieve the project objectives on time and to the specified cost, quality and performance. (Project Management nstitute - PM) Resource Scheduling Project scheduling with infinite capacity Resources allocated profile evaluation omparison between resources requirements / availability Resource leveling Resource constrained project scheduling problem (RPSP): fixed times or resources Mauro Mancini Mauro Mancini

2 Resource Scheduling PM vs. Resource Scheduling Reporting Project disaggregation in more activities Precedence relationships among activities Trade - off Length activities/ resources allocation Project s resources reallocation PM Resources requirements / availability evaluation Project length assessment Plan execution PM onsider infinite capacity Precedence are stable relationships ritical chain is a function of time and natural precedence Activities floats depend only on project duration Resource Scheduling onsider limits on resources availability mbodying all the kinds of relationships, consider that they may be dynamic and temporary ritical chain is a function of time, all the kinds of relationships, resources need and availability, access priority of the activities to the resources Activities floats depend on resources need and availability Mauro Mancini Mauro Mancini Steps to apply a Resource Management methodology. nfinite capacity programming. Aggregate resource need evaluation. umulative resources need evaluation. inite capacity programming (Resource onstrained Project Scheduling Problem). Levelling resources need profile (Resource Leveling) Resources Overload Schedulation Types Time Resources Resources Levelling Limitation Time ixed time ixed resources Mauro Mancini Time Mauro Mancini

3 Resource Levelling Resource leveling aims to minimize the period-by-period variations in resource loading by shifting tasks within their slack allowances. The purpose is to create a smoother distribution of resource usage. Advantages: much less hands-on management is required if the use of a given resource is nearly constant over its period of use. The PM can arrange to have the resource available when needed can have the supplier furnish constant amounts can arrange for a backup supplier if advisable resource usage is leveled, the PM may be able to use a just-intime inventory policy without much worry that the quantity delivered will be wrong Resource Levelling advantages when resources are leveled, the associated costs also tend to be leveled it is a procedure that can be used for almost all projects, whether or not resources are constrained Mauro Mancini Mauro Mancini 0 Resource onstrained Project Scheduling problem (RPSP) Resource onstrained Project Scheduling problem (RPSP) Optimization techniques Limits: heavy computation static definition of the problem difficulty to determine optimality criteria, considering several possible objectives in contrast. euristic approaches Limits: they can pursue only one objective for every scheduling they obtain different performance according to the network characteristics to which they are applied they are based on priority rules (to solve conflict between activities that require common resources) fixed a priori, without the possibility to adapt them to the project characteristics. Mauro Mancini Mauro Mancini

4 Resource onstrained Project Scheduling problem (RPSP) euristic approaches valuation of conflicts among activities to the access to shared resources One time and at local level n each step: RPSP Determination of the activities list with precedence constraints verified (in competition with shared resources) Activities group selection with the resource availability having: ominimum delay of project completion opriority rules Parallel approach, forward, early start (front loading) Mauro Mancini Mauro Mancini euristic Approach RPSP n series approach (static) oordinate list of starting activities ot tends to advance by paths, in order of criticality n parallel approach (dynamic) oordinate list of activities updated at each step ot tends to advance by temporal intervals RPSP n parallel approach Rolling wave approach Different priority rules (problem s dependent) Splitting / interrupting Multi-resources vision What if analysis Mauro Mancini Mauro Mancini

5 RPSP RPSP OW TO NRAS RSOURS Working Overtime Working Shifts ncrease Productivity Job and Knock Learning urve Sub-ontractors Scope of Work OW TO RDU RSOURS Move unemployed resources to other activities Move unemployed resources to R&D jobs ire out resources internally or externally Pre-manufacture components before they are needed Maintenance of equipment during slack periods Train workforce during slack periods to gain new skills which will make them more productive and flexible in the future Send the under utilized workforce on leave Mauro Mancini Mauro Mancini Series Parallel Approach P P P Priority rules MN T (Minimum Total loat) The conflict to access to shared resources by different activities is solved giving priority to the activity with minimum float ritical Path Series approach Parallel approach t s possible demonstrate that, using a parallel approach, the MN T is equivalent to the MN LST rule that grades activities according to the increasing values from the starting time at the latest Mauro Mancini Mauro Mancini 0

6 Priority rules RSM ( Resource Scheduling Method ) The precedence is given to the activity characterized by the minimum value for the d ij parameter. t corresponds to the project duration increment that occurs when the j activity follows the i activity with: d ij = max [0; T i -LST j ] Where: T i = arliest finish time for the i activity LST j = Latest start time for the j activity Priority rules RSM ( Resource Scheduling Method ) The comparison is made among all the pairs belonging to the sets of activities in conflict for the access to shared resources. This rule gives results similar to the MN LT rule. LST i d ij T i MN LT ( Minimum Late inish Time ) This rule sequences the activities giving precedence to activities characterized by minimum late finish Time. Mauro Mancini Mauro Mancini Priority rules GRD ( Greatest Resource Demand ) t use the priority criterion of unit s resource requirements (considering all the demanded resources) associated to each activity, giving precedence to the activities characterized by a greater requirements (they can potentially create bottlenecks). The priority grade associated to each activity is computed in this way: Priority = d i Σ r ij i = Where: dj = j activity duration rij = i type resource requirements for each unit of time by the j activity m = number of the different types of resources m Priority rules GRU ( Greatest Resource Utilization ) This rule gives the priority, in each program step, to the combination of activities that obtaining the maximal possible saturation for the available resources. The rule implementation requires the resolution of a linear programming problem of integer number (0,). SO ( Shortest mminent Operation ) The priority is given to the activities with shortest duration. This rule comes from job-shop scheduling problem. Mauro Mancini Mauro Mancini

7 Priority rules MJP ( Most Job Possible ) This rule gives the priority to the combination of activities that allows to program the maximum possible number of activities in each programming step. t differs to the GRU rule because the determination of the maximum possible number of activities considers only the feasibility, that is the availability of resources, and not their saturation level. Series Parallel Approach xercise onsidering a project composed by the following activities A D L 0 Att. Ra Dur. Rb N 0 M 0 G 0 0 Mauro Mancini Mauro Mancini Series Parallel Approach xercise. To compute the ST, T, LST, LT and the resource allocated temporal profile. Using the min total float with fixed time levelling and resource availability of units/day for the A resource (and infinite availability for the resource) to schedule the activities with: Series approach Parallel approach. Like at point, but levelling with fixed resources respecting the availability for the A and resources ( units/day) Mauro Mancini Series Parallel Approach xercise rom the provided data is possible to obtain: A 0 0 G L 0 D ST LST T A 0 Mauro Mancini M 0 LT N 0

8 Series Parallel Approach xercise Attività ST T LST LT T Ra Rb A D G L - M - - N - - Series Parallel Approach xercise A D G L M N 0 Mauro Mancini Mauro Mancini 0 Series Parallel Approach xercise Series Parallel Approach xercise 0 Resource - A 0 0 Resource - 0 Mauro Mancini Mauro Mancini

9 Series Approach ixed Times. The activities are scheduled considering the min total float Series Approach ixed Times. The activities are scheduled considering the min total float Activity M N D L A G T The activities are scheduled considering the network diagram Available AG Available AGD Available AGD D Activity M N D L A G T The activities are scheduled considering the network diagram Available AG Available AGD Available AGD D Scheduling order 0 D A L G M N Mauro Mancini Mauro Mancini Series Approach ixed Times Series Approach ixed Times Programming phase (fixed times) A The activity has been programmed (days - with days delay resp. A) The activity has been programmed (days -0 with days delay resp. A) The D activity has been programmed (days - with 0 days delay resp. A) D The predetermined order is closely followed G L M N 0 Mauro Mancini Mauro Mancini

10 Series Approach ixed Times Parallel Approach ixed Times 0 Resource A (series) M A G M N 0 Algorithm s step:. reation the list of programmable activities. Re-compute the relative T and ST. Tidy up the list activities. ollocated one at a time following the point order. ome back to point n order to respect the project deadline is necessary to have an overload () Mauro Mancini Mauro Mancini Parallel Approach ixed Times Parallel Approach ixed Times Step DAT AVALAL ATVTS TOTAL LOAT RSUR A RQURMNTS Step DAT AVALAL ATVTS UPDATD TOTAL LOAT RSUR A RQURMNTS Step 0 d. A DAT G AVALAL ATVTS UPDATD TOTAL LOAT RSUR A RQURMNTS 0 d. D - A G Step G d. - DAT AVALAL ATVTS UPDATD TOTAL LOAT RSUR A RQURMNTS d. The available resource is not enough, but is necessary starting the activity the day, in fact in order to respect the resource constraint, should be postpone to the day, with days project delay, D, A start. G has been postponed because of lack of resources Mauro Mancini Mauro Mancini 0

11 Parallel Approach ixed Times Parallel Approach ixed Times Step DAT AVALAL ATVTS UPDATD TOTAL LOAT RSUR A RQURMNTS Step DAT AVALAL ATVTS UPDATD TOTAL LOAT RSUR A RQURMNTS d. 0 / d. L 0 - Step DAT AVALAL ATVTS UPDATD TOTAL LOAT RSUR A RQURMNTS Step DAT AVALAL ATVTS UPDATD TOTAL LOAT RSUR A RQURMNTS d. M 0 d. N 0, D, A start. G has been postponed because of lack of resources Mauro Mancini Mauro Mancini Parallel Approach ixed Times Parallel Approach ixed Times A D G L M N 0 0 Resource A (Parallel) A G M M N 0 Mauro Mancini Mauro Mancini

12 Series and Parallel Approach ixed Resources. n case of fixed resources programming the project deadline constraint is not considered, but it is no longer possible having overload. The series and parallel approach stay unchanged. The same schedule remains or the series approach. Mauro Mancini Parallel Approach ixed Times Resource - A A G M N 0 0 Resource - D L 0 Mauro Mancini Series Approach ixed Resources Series Approach ixed Resources A D L A D G L G M N M N Mauro Mancini Mauro Mancini

13 Parallel Approach ixed Resources Suggestions Resource A A G M N 0 0 Resource - D L 0 To obtain the best results it is convenient to adopt the following expedient: to use preferably parallel approach to attempt different priority rules to execute resource leveling along a limited temporal period (according to a rolling wave approach) to use options available on the software (splitting, interrupting, reprofiling, etc.) to verify the possibility to maintain a multi-resource vision of the utilization resources profile coming from scheduling to led what-if analysis, hypothesizing possible corrective actions Mauro Mancini Mauro Mancini 0 xample The Gantt and precedence diagram Activity Duration Precendence Resource A g a[0] g A b[0] g a[0] D g b[0] g a[0] g D a[0] G g a[0] a[0];b[0 g ] g G a[0] L g ; b[0] M g ; a[0] N g L;M a[0] ind the critical path, and the resources utilisation diagram (hypothesis of infinite resources available). Assume to start on september D Nome attività Durata nizio ine Predeces Nomi risorse A g lun 0/0/0 mar 0/0/0 a[0] g mer 0/0/0 sab 0/0/0 b[0] g lun 0/0/0 mar 0/0/0 a[0] D g mer 0/0/0 dom 0/0/0 b[0] g dom 0/0/0 dom 0/0/0 a[0] g lun 0/0/0 gio /0/0 a[0] G g lun 0/0/0 lun 0/0/0 a[0] g mer 0/0/0 mer 0/0/0 a[0];b[0] g mar 0/0/0 gio 0/0/0 a[0] 0 L g ven /0/0 dom /0/0 ; b[0] M g gio /0/0 lun /0/0 ; a[0] N g mar /0/0 mer /0/0 0; a[0] 0 set 0 0 set 0 set 0 L M M G V S D L M M G V S D L M M G V S D Mauro Mancini Mauro Mancini

14 Resource a: utilisation diagram Resource b: utilisation diagram Mauro Mancini Mauro Mancini Resource constraints Resource constraints: Assume now that only resources a are available. What s happen? Solve the problem with a series approach Mauro Mancini Mauro Mancini

15 Schedule: step Acivity duration ST T LST LT T Ra Rb A 0 D G L M 0 N 0 Mauro Mancini Schedule: step Schedule with the serial approach 0 D A L G M N With same project duration resources are not enough (even with a parallel approach) Mauro Mancini A possible solution: to add more days A possible solution: to add more days D Nome attività Durata nizio ine Predeces Nomi risorse A g mer 0/0/0 gio 0/0/0 a[0] g v en 0/0/0 lun 0/0/0 b[0] g lun 0/0/0 mar 0/0/0 a[0] D g mer 0/0/0 dom 0/0/0 b[0] g lun /0/0 lun /0/0 a[0] g lun 0/0/0 gio /0/0 a[0] G g mer 0/0/0 mer 0/0/0 a[0] g mer 0/0/0 gio /0/0 a[0];b[0] g v en /0/0 dom /0/0 a[0] 0 L g mar /0/0 gio /0/0 ; b[0] M g lun /0/0 v en /0/0 ; a[0] N g sab 0/0/0 dom /0/0 0; a[0] 0 set 0 0 set 0 set 0 set L M M L M M L M M L M G V S D G V S D G V S D t was the /0 Mauro Mancini Mauro Mancini 0