1. An Architecture for Scheduling and Control in Flexible Manufacturing Systems Using Distributed Objects TsuTa Tai and Thomas O. Boucher Presented by: Ammon Johnson November 10, 2008

2. Function of Paper Use a decentralized approach to solve scheduling problems Use a decentralized approach to solve scheduling problems Optimize scheduling when changes happen in the system Optimize scheduling when changes happen in the system Deadlock avoidance Deadlock avoidance Compare effectiveness and computation time Compare effectiveness and computation time

3. Importance Reducing deadlock, time to manufacture (makespan) will improve profitability of the manufacturing operation Reducing deadlock, time to manufacture (makespan) will improve profitability of the manufacturing operation Scheduling is dynamic; Sudden changes can adversely affect productivity Scheduling is dynamic; Sudden changes can adversely affect productivity

4. References REFERENCES [1] N. Costa and M. Garetti, “Design of a control system for a flexible manufacturing cell,” J. Manuf. Syst., vol. 4, pp. 65–84, 1984. [2] T. O. Boucher,M. A. Jafari, and G. A. Meredith, “Petri net control of an automated manufacturing cell,” Adv. Manuf. Eng., vol. 2, pp. 151–157, 1990. [3] H. P. Huang and P. C. Chang, “Specification, modeling and control of a flexible manufacturing cell,” Int. J. Prod. Res., vol. 30, pp. 2515–2543, 1992. [4] S. B. Joshi, E. G. Mettala, J. S. Smith, and R. A.Wysk, “Formal models for control of flexible manufacturing cells: Physical and system models,” IEEE Trans. Robot. Automat., vol. 11, pp. 558–570, Aug. 1995. [5] A.Yalcin and T. O. Boucher, “An architecture for flexible manufacturing cells with alternate machining and alternate sequencing,” IEEE Trans. Robot. Automat., vol. 15, pp. 1126–1130, Dec. 1999. [6] N. Viswandham, Y. Narahari, and T. L. Johnson, “Deadlock prevention and deadlock avoidance in flexible manufacturing systems using Petri net models,” IEEE Trans. Robot. Automat., vol. 6, pp. 713–723, Dec. 1990. [7] J. Ezpeleta, J. M. Colom, and J. Martinez, “A Petri net-based deadlock prevention policy for flexible manufacturing systems,” IEEE Trans. Robot. Automat., vol. 11, pp. 173–184, Apr. 1995. [8] J. Ezpeleta and J. M. Colom, “Automatic synthesis of colored Petri nets for control of FMS,” IEEE Trans. Robot. Automat., vol. 13, pp. 327–337, June 1997.

5. References (cont.) [9] R. A.Wysk, N. S. Yang, and S. Joshi, “Detection of deadlocks in flexible manufacturing cells,” IEEE Trans. Robot. Automat., vol. 7, pp. 853–859, Dec. 1991. [10], “Resolution of deadlocks in flexible manufacturing systems: Avoidance and recovery approaches,” J. Manuf. Syst., vol. 13, pp. 128–138, 1999. [11] Z. A. Banaszak and B. H. Krogh, “Deadlock avoidance in flexible manufacturing systems with concurrently competing process flows,” IEEE Trans. Robot. Automat., vol. 6, pp. 724–734, Dec. 1990. [12] F. S. Hsieh and S. C. Chang, “Dispatching-driven deadlock avoidance controller synthesis for flexible manufacturing systems,” IEEE Trans. Robot. Automat., vol. 10, pp. 196–209, Apr. 1994. [13] K. Y. Xing, B. S. Hu, and H. X. Chen, “Deadlock avoidance policy for Petri net modeling of flexible manufacturing systems with shared resources,” IEEE Trans. Automat. Contr., vol. 41, pp. 289–295, Feb. 1996. [14] M. P. Fanti, B. Maione, S. Mascolo, and B. Turchiano, “Event-based feedback control for deadlock avoidance in flexible production systems,” IEEE Trans. Robot. Automat., vol. 13, pp. 347–363, June 1997. [15] M. A. Lawley, S. A. Reveliotis, and P. M. Ferreira, “A correct and scaleable deadlock avoidance policy for flexible manufacturing systems,” IEEE Trans. Robot. Automat., vol. 14, pp. 796–809, Oct. 1998. [16] N. Q. Wu, “Necessary and sufficient conditions for deadlock-free operation in flexible manufacturing system using colored Petri net model,” IEEE Trans. Syst., Man, Cybern., vol. 29, pp. 192–204, May 1999. [17] M. A. Lawley, “Deadlock avoidance for production systems with flexible routing,” IEEE Trans. Robot. Automat., vol. 15, pp. 497–509, June 1999.

6. References (cont.) [18] A. Yalcin and T. O. Boucher, “Deadlock avoidance in flexible manufacturing systems using finite automata,” IEEE Trans. Robot. Automat., vol. 16, pp. 424–429, Aug. 2000. [19] T. O. Boucher, A. Yalcin, and T. Tai, “Dynamic routing and the performance of automated manufacturing cells,” IIE Trans., vol. 32, no. 10, pp. 975–988, 2000. [20] D. Y. Lee and F. DiCesare, “Scheduling flexible manufacturing systems using Petri nets and heuristic search,” IEEE Trans. Robot. Automat., vol. 10, pp. 123–132, Apr. 1994. [21] S. E. Ramaswamy and S. B. Joshi, “Deadlock-free schedules for automated manufacturing workstations,” IEEE Trans. Robot. Automat., vol. 12, pp. 391–400, June 1996. [22] H. H. Xiong and M. C. Zhou, “A Petri net method for deadlock-free scheduling of flexible manufacturing systems,” Int. J. Intell. Contr. Syst., vol. 3, pp. 277–295, 1999. [23] J. Pearl, Heuristics: Intelligent Search Strategies for Computer Problem Solving. Reading, MA: Addison-Wesley, 1984. [24] R. E. Tarjan, “Depth first search and linear graph algorithm,” SIAM J. Comput., vol. 1, pp. 146–160, 1972. [25] A. Yalcin, “Architectures for automated flexible manufacturing cells with routing flexibility,” Ph.D. dissertation, Rutgers Univ., New Brunswick, NJ, 2000. [26] J. Ferber, Multi-Agent Systems. Reading, MA: Addison-Wesley, 1999. [27] R. Smith, “The contract net protocol: High-level communication and control in a distributed problem solver,” IEEE Trans. Comput., vol. C–23, pp. 1104–1113, 1980.

7. References (cont.) [28] R. Smith and R. Davis, “Framework for co-operation in distributed problem solving,” IEEE Trans. Syst., Man, Cybern., vol. SMC–11, pp. 61–70, 1981. [29] W. D. Kelton, R. P. Sadowski, and D. A. Sadowski, Simulation With Arena. New York: McGraw-Hill, 1998. [30] D. C. Montgomery, Design and Analysis of Experiments. New York: Wiley, 1976. [31] T. Tai and T. O. Boucher, “Scheduling With Distributed Objects: Source Code and Experimental Trials,” Ind. Eng. Dept., Rutgers Univ., Piscataway, NJ, Working Paper #01-119, 2001.

8. Relation to ME 482 Scheduling is one of the most complex aspects of FMS Scheduling is one of the most complex aspects of FMS Optimizing the scheduling of tasks is a tedious task, so computers are used to optimize the scheduling Optimizing the scheduling of tasks is a tedious task, so computers are used to optimize the scheduling

9. Basic Design Concept Shop Floor Object (Central control computer) Cell Object

10. Basic Design Concept Shop Floor Object (Central control computer) Cell Object New Part Determines cell with shortest “makespan” Part goes to cell with shortest makespan

11. Design Principle Algorithm development Algorithm development DFS (Depth First Search)- looks for end DFS (Depth First Search)- looks for end DFS with Greedy Heuristic DFS with Greedy Heuristic DFS Greedy with Knot Detection DFS Greedy with Knot Detection

12. Process plan and digraphs for one cell object In this example the cells algorithm generates a legal sequence of events, and avoids deadlock to finish both parts. May be more than one legal sequence. Cell object generates a schedule that finishes all current parts and the new part.

13. Digraphs DFS Greedy Heuristic DFS Greedy Heuristic DFS Greedy with DFS Greedy with Knot detection

14. Example Problem Time for processing parts A and BDeadlock free schedule

15. Experimental Trials Four different scheduling rules Four different scheduling rules Queue length (Q)- part goes to shortest line Queue length (Q)- part goes to shortest line Bottleneck machining time (BT)- cell with least additional bottleneck time added Bottleneck machining time (BT)- cell with least additional bottleneck time added Balanced workload scheduling (BL) Balanced workload scheduling (BL) These three were compared to the distributed object method (DO) These three were compared to the distributed object method (DO) Three cell system Three cell system Ten paired comparisons Ten paired comparisons of 100 parts each

16. Experimental Equipment The equipment consists of different software modules used to simulate the factory environment The equipment consists of different software modules used to simulate the factory environment Simulator sends a “new part” to the cell, receives makespans, and assigns the part Simulator sends a “new part” to the cell, receives makespans, and assigns the part

17. Experimental Results Algorithms discussed were applied Algorithms discussed were applied Distributed object scheduling compared with Q, BL, and BT Distributed object scheduling compared with Q, BL, and BT

18. Experimental Results Average makespan 9-14% lower than other methods Average makespan 9-14% lower than other methods Throughput is increased Throughput is increased Computation is very fast Computation is very fast

19. Correlation of Results with Model Authors are unsure of source of improvement in performance Authors are unsure of source of improvement in performance Not a very complex system (3 cells) Not a very complex system (3 cells) The simulation is both model and experiment The simulation is both model and experiment

20. Practical Industrial Use and Advancement Shows that throughput was increased, makespan decreased in simulation Shows that throughput was increased, makespan decreased in simulation No comparison with actual hardware No comparison with actual hardware Advancement in scheduling FMS, improving production Advancement in scheduling FMS, improving production Industries that use FMS systems, auto, aerospace, etc. Industries that use FMS systems, auto, aerospace, etc.