ISPE Workshop Hamburg, April 2013 Regulatorischer Druck und Kostendruck in der Pharmaindustrie: Trends zur Veränderung der Produktionsprozesse

Transcription

1 ISPE Workshop Hamburg, April 2013 Regulatorischer Druck und Kostendruck in der Pharmaindustrie: Trends zur Veränderung der Produktionsprozesse Trend 8: durch einen modellbasierenden Ansatz: ein Weg zu besserer Kapitaleffizienz, höherer Geschwindigkeit, Kostenreduzierung und dynamischer Adaptierbarkeit. Ref. M. Janovjak, 15th April

2 durch einen modellbasierenden Ansatz Content 1. Challenge 2. Systemic view 3. Model based approach 4. Methodology & Tool 4. Applications (Examples) 5. Conclusions Ref. M. Janovjak, 15th April

4 durch einen modellbasierenden Ansatz, 1. Challenge The complexity of Manufacturing supply chain processes often overwhelms our ability to understand them. In this perspective we are challenged to evolve an advanced approach to master it. Ref. M. Janovjak, 15th April

5 durch einen modellbasierenden Ansatz, 1. Challenge An advanced approach is needed that enables users - to handle dynamic complexity (timely changes of e.g. product portfolio, disposition of operational processes, trade offs ) - to optimize performance (costs reduction, maximization throughput, OEE, cycle time, process capability ) - to utilize environment (small order sizes, responsiveness, resources, shift / time regime ) at the same time in a balanced and pro-active way Ref. M. Janovjak, 15th April

6 durch einen modellbasierenden Ansatz, 1. Challenge Such advanced approach - requires a systemic approach, which concentrates on the whole reference system rather than on segments or single events. - calls for understanding the causal structures and generative mechanisms of the manufacturing supply chain which govern its system behavior. - must be based on a dynamic rather than static view (analytical and synthetic). Ref. M. Janovjak, 15th April

8 System durch einen modellbasierenden Ansatz, 2. Systemic view KPI, Benefit potentials demand Manufacturing SC process supply Levers Ref. M. Janovjak, 15th April

9 System System elements durch einen modellbasierenden Ansatz, 2. Systemic view KPI, Benefit potentials Optimization of assets & capacity utilization; Reduce inventory levels; Reduce Cycle Time; Maximize throughput; Improve Process capability & robustness; demand Process Manufacturing steps & structure; Equipment characterization; SC Process parameters; Materials; process Resources; supply Process design; Policies and principles; Dynamic & linkages of processes; Product portfolio & product quality; Demand (order sizes, scheduling, ) Levers Ref. M. Janovjak, 15th April

10 System durch einen modellbasierenden Ansatz, 2. Systemic view KPI, Benefit potentials Manufacturing SC Process schematics demand Process parameters Equipment characterization Manufacturing Process steps SC & structure process supply Materials Resources Levers Ref. M. Janovjak, 15th April

12 durch einen modellbasierenden Ansatz, 3. Model based approach How to evolve Model based approach Build a model of the real system and provide simulation for determination of the best solution. real system Such Model based approach requires - dynamic and causal process understanding - consideration of data and expertise - implementation of policies and principles - application of adequate methodology Ref. M. Janovjak, 15th April

13 durch einen modellbasierenden Ansatz, 3. Model based approach Methodology Tool real system Expertise Policies Principles Modelling data, structure, interactions Model Optimization cycle Simulation results Implement best solution Ref. M. Janovjak, 15th April

14 durch einen modellbasierenden Ansatz, 3. Model based approach Model is simplified but validated representations of the real system. Model consists all relevant cause-and-effect relationships of the manufacturing SC System components, incl. feedbacks Simulation explores the sensitivity of the model behaviour and based on it determines any alternatives and what-if scenarios of the answer to a reel problem. The simultaneous and complementary (multifunctional) consideration of levers and KPI s enables to optimize the manufacturing SC Process and to determine the best performing solution. Ref. M. Janovjak, 15th April

15 durch einen modellbasierenden Ansatz, 3. Model based approach The Model based approach is one and only way to provide solutions in pro-active way. The capability to act pro-active opens up additional value proposition - process design - capital efficiency - implementation speed - cost avoidance - risk mitigation On the purpose, the capability of the applied methodology and tool (SW) has to be of unique performance. Ref. M. Janovjak, 15th April

17 durch einen modellbasierenden Ansatz, 4. Methodology & Tool System Dynamics is a holistic universal applicable Methodology for studying & managing complex feedback systems.. Vensim is a tool (SW) applying SD methodology which enables - to transform knowledge and data into model - to build a model of any complex system - to determine the best performing solution under consideration of the multidimensional optimization policy. Ref. M. Janovjak, 15th April

18 durch einen modellbasierenden Ansatz, 3. Model based approach Vensim Models delivers (SOURCE: VENTANA LTD). Intelligence amid mixed signals by combining human expertise and data to determine the true signal Focus amid distractions by using cumulative knowledge to highlight new leverage Consensus amid multiple views by integrating knowledge and reconciling differences Understanding amid complexity By tracking the interactions of causes and effects,... inventory control production rate production inventory production status change over Confidence amid uncertainty by showing the odds and the best ways to influence them while experience deepens Ref. M. Janovjak, 15th April

20 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Depending on focus (Process, Purpose, Objectives) there is a wide spectrum of possible applications, exemplary: Manufacturing: demand and supply, material & product flows, demand & order tracking, WIP / Inventory, throughput, Cycle Time, yield, utilization / OEE, Resources Manufacturing: processing & technology process parameter variations and associated risks, impact on product quality Process development speed / time to market, cost reduction Product Portfolio Life Cycle Management time to market, costs, risks, value generation Ref. M. Janovjak, 15th April

21 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Manufacturing: demand and supply, material & product flows, Solid manufacturing & packaging (coated tablets, Lean / WIP, Cycle Time, ) Biopharmaceuticals bulk manufacturing & fill & finish (organizational design /WIP, CT, ) Liquid fill & finish (syrup, Design & investment efficiency) Biopharmaceuticals bulk manufacturing & fill & finish (PFS, Lean / WIP, SC responsiveness, ) Packaging & Palletizing area (demand & performance variations, Eq. design, ) Ref. M. Janovjak, 15th April

22 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Manufacturing: processing & technology Process model and risk based multivariate quantitative investigational analysis and improvement of an API process consisting multiple process steps (Dialysis, Ultra filtration, Distillation, Sterile Filtration) Process development Process development and transfer (solid, granulation) Product Portfolio Life Cycle management PPLCM Management and strategic governance (LC Phases II to launch (Pilot Study)) For more information about the 3 Ex. above, please contact InSysteA Ref. M. Janovjak, 15th April

24 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Solid manufacturing & packaging Purpose Build a model which enables to elaborate the Lean optimized manufacturing and packaging process (one certain coated tablet product) Starting position Order based packaging installed, rhythm wheel optimized push practice in bulk manufacturing (at each work center) Approach Objectives Deployment of Pull principle across entire process, optimized campaign size in dispensing & granulation, batch and customer order tracing, back log control, representation of QC & QA process, Eq. utilization Optimization Objectives CT reduction (responsiveness), Cost Reduction (WIP, Inventories, batch yield) Results Cost reduction through reduction of WIP / Inventory ca. 300 k CHF/a while maintaining same customer service KPI Ref. M. Janovjak, 15th April

25 20,000 15,000 10,000 5, Day SimulatedPullSystem - Dispensing SimulatedPullSystem - Granulating SimulatedPullSystem - Tabletting SimulatedPullSystem - Coating SimulatedPullSystem - Packing InSysteA durch einen modellbasierenden Ansatz, 5. Applications (Examples) Solid manufacturing & packaging Model Parameterization Initialization Pre-position batches Dispensing Granulation Tabletting Coating Packaging Pull Capacity / work center optimized Order rhythm Pull wheel Principle push across practice entire process based Warehouse Backlog control per each process step Dispensing Granulation Tabletting Coating Scheduling supply demand Batch tracing, QC & QA processing Customer order tracing Packaging mg Material Losses (Stacked) mg Material Losses (Stacked) Total Product at Location (Stacked) Total wait time - 25mg (Stacked) Total wait time - 50mg (Stacked) 100 Machine Utilisation - (running average) day (Day) SimulatedPullSystem - Dispensing SimulatedPullSystem - Granulating SimulatedPullSystem - Tabletting SimulatedPullSystem - Coating SimulatedPullSystem - Packing mg Material Losses (Stacked) day (Day) SimulatedPullSystem - Dispensing SimulatedPullSystem - Granulating SimulatedPullSystem - Tabletting SimulatedPullSystem - Coating SimulatedPullSystem - Packing day (Day) SimulatedPullSystem - Dispensing SimulatedPullSystem - Granulating SimulatedPullSystem - Tabletting SimulatedPullSystem - Coating SimulatedPullSystem - Packing Yield losses 200 mg Material Losses (Stacked) day (Day) SimulatedPullSystem - Dispensing SimulatedPullSystem - Granulating SimulatedPullSystem - Tabletting SimulatedPullSystem - Coating SimulatedPullSystem - Packing 1, WIP mfct. Mean Cycle Times (Stacked) Ref. M. Janovjak, 15th April day (Day) SimulatedPullSystem - Dispensing Hour SimulatedPullSystem - Granulating Hour SimulatedPullSystem - Tabletting Hour SimulatedPullSystem - Coating Hour Total wait time - 100mg (Stacked) day (Day) Cycle Times day (Day) SimulatedPullSystem - Dispensing Hour Dispensing Hour SimulatedPullSystem - Granulating Hour Granulating Hour SimulatedPullSystem - Tabletting Hour Tabletting Hour SimulatedPullSystem - Coating Hour Coating Hour Idle times day (Day) SimulatedPullSystem - Dispensing Hour SimulatedPullSystem - Granulating Hour SimulatedPullSystem - Tabletting Hour SimulatedPullSystem - Coating Hour Total wait time - 200mg (Stacked) day (Day) SimulatedPullSystem - Dispensing Hour SimulatedPullSystem - Granulating Hour SimulatedPullSystem - Tabletting Hour SimulatedPullSystem - Coating Hour day (Day) SimulatedPullSystem -Granulating SimulatedPullSystem -Tabletting SimulatedPullSystem -Coating SimulatedPullSystem -Packaging Eq. utilization Percent Percent Percent Percent Simulation 30 days; 90 orders;

27 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Biopharmaceuticals bulk manufacturing & fill & finish Source: Lee Jones, Ventana Ltd. Purpose Assess impact of change from functional to process-oriented organization given; Random demand stream for multiple products Shared production lines Alternate physical locations of packaging Approach Study current system and build as is SC model; validate Identify performance and structural changes when moving from functional to process-oriented organization ( to be ) Modify model/parameter values to represent multiple to be options Evaluate impact on KPI such as delivery performance, stock holding and cycle times, and determine possible benefit potentials Results Reduction in cycle time of (~25% ) and stocks (~33%) while maintaining same customer service KPI Ref. M. Janovjak, 15th April

28 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Biopharmaceuticals bulk manufacturing & fill & finish Source: Lee Jones, Ventana Ltd. Order data and informs WIP Informs forecasting scheduling Orders and order tracing drives order data Production campaign drives by pulled forecast scheduling WIP production campaign WIP forecast Orders packaging schedule Pull bulk for packaging Orders drives packaging schedule Order tracing Product pushed though packaging to finished goods and delivery thawing Pull API through production compounding filling transport bulk storage Product pushed pulled through filling process Primary contained product WIP transport packaging transport finished delivery goods Materials OEE FTE WIP Availabilility of FTE and Materials, packaging line capability influence packaging performance Ref. M. Janovjak, 15th April

30 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Liquid fill & finish Purpose Assess the costs reduction potential of currently coupled line and determine the possible benefits in the case of decoupling of the line. Starting position Order based SC installed, optimized rhythm wheel controlled manufacturing ( push ); planned project for line decoupling; Approach Objectives Determination of the solution with biggest cost reduction potential, incl. decision support for project realization. Lean optimization, batch and customer order tracing, representation of QC & QA process, Eq. utilization; Optimization Objectives Cost Reduction (WIP / Inventories, change over costs) Results Cost reduction Lean optimized current coupled line: 25 % Cost reduction Lean optimized decoupled line (planned project): < 30 % Planned project cancelled due to insufficient rentability (cost avoidance 250 kchf) Ref. M. Janovjak, 15th April

31 Coupled line - Rhythm wheel Push Coupled line - Pull (lean optimized current) Decoupled line - Push on Fill / Pull on Pack Decoupled line - Pull on Fill&Pack Decoupled line - Pull on Fill&Pack (QC parallel) InSysteA durch einen modellbasierenden Ansatz, 5. Applications (Examples) Liquid fill & finish Costs (%) Filling Packaging Total Costs Coupled line 110% Decoupled line (planned project) Inventory Change over Fill + Pack Fill Pack (Simulation period = 500 days) Coupled Decoupled line line - - Coupled Decoupled line line 30 - Rhythm - line Pull - (lean Push wheel P Coupled Decoupled line line - Rhythm 25 - Push Pull % line (lean on - wheel Pull Fill optimized on / Push Pull potential Fill&Pack on current) < 30 % Coupled Decoupled line - Pull line (lean - 100% 90% 80% 70% 60% 50% Cost reduction Cost reduction Coupled line - Rhythm wheel Push (current) Coupled line - Pull (lean optimized ) Decoupled line - Push on Fill ; Pull on Pack Decoupled line - Pull on Fill & Pack Policy optimized Decoupled line - Pull on Fill & Pack and QC parallel Ref. M. Janovjak, 15th April

33 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Biopharmaceuticals bulk manufacturing & fill & finish Purpose Build a model which enables to elaborate the Lean optimized manufacturing SC process Starting position Order based SC installed, BB Cycle Time focused project implemented Approach Objectives Deployment of Pull principle, optimized batch and customer order tracing, back log control, representation of QC & QA process, Eq. utilization Optimization Objectives CT reduction (responsiveness), Cost Reduction (WIP, Inventories) Results Cost reduction through reduction of 4 batches of 24 in WIP Inventory cost reduction ca k CHF/a Ref. M. Janovjak, 15th April

34 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Biopharmaceuticals bulk manufacturing & fill & finish Planned batch orders Raw materials Frozen API Additives I Additives II Needle Shields PFS barrel Stoppers Primary Packaging Preparation and assembling Simulation period 60 days; 469 orders Model Parameterization batch tracing first In line filters Initialization Pre-position batches Compounding and Filling line preparation Pull principle customer order tracing compounding filling optical inspection packaging Subscripts Calculations results : Results : Scheduling - Cycle times at each process steps - mean Cycle times stacked - Total product demand & available supply (released filled product prior sec. packaging) - customer order backlogs (duration, # of orders, at delivery) - stock levels WIP at product and at process step (Lt., # of syringes) - Yield / CONC - Equipment utilization (OEE) Ref. M. Janovjak, 15th April

35 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Manufacturing: demand and supply, material & product flows, Solid manufacturing & packaging (coated tablets, Lean / WIP, Cycle Time, ) Biopharmaceuticals bulk manufacturing & fill & finish (organizational design /WIP, CT, ) Liquid fill & finish (syrup, Design & investment efficiency) Biopharmaceuticals fill & finish (PFS, Lean / WIP, SC responsiveness, ) Packaging & Palletizing area (demand & performance variations, Eq. design, ) Ref. M. Janovjak, 15th April

36 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Packaging & Palletizing area Purpose Determine performance of packaging area and needed size of conveyor system and palletizer and provide sensitivity analysis of impact of variation of order sizes and lines performance on results. Starting position Upgrade of existing production. Approach Objectives Deployment of random functions to simulate variations of - order sizes - lines performance - change over and line clearance Optimization Objectives & Results - Determination of impact of the shift to smaller orders sizes on performance and throughput - Determination of material flow (cases) and palettizer utilization (mean, St. Dev.) - Model design adaptable to extended scope (scheduling, order control, WIP ) Ref. M. Janovjak, 15th April

37 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Packaging & Palletizing area Packaging line outflow in cases Packaging line outflow accumulated in cases, conveyor system & transport to Palletizer Palletizing process as flow rate into buffer in front of Palletizer and warehouse Palletizer performance & utilization Randomized order size and line performance and determination of batch duration and processing time Batch control cycle with determination of cumulative # of batches and change order execution data Randomized duration change order and share fraction based decision of change order activity Ref. M. Janovjak, 15th April

38 durch einen modellbasierenden Ansatz, 5. Applications (Examples) Packaging & Palletizing area P40 P50 P45 max ,000 Palletizer Buffer 60 Palletizer outflow per hour 4, VVP 3,000 Pallet/Hour 30 1, max palletising Time (Minute) Palletizer Buffer : P40 Palletizer Buffer : P50 Palletizer Buffer : P Time (Minute) Palletizer outflow per hour : P40 Palletizer outflow per hour : P50 Palletizer outflow per hour : P45 Simulation period 30 days; orders Ref. M. Janovjak, 15th April

39 durch einen modellbasierenden Ansatz, 5. Conclusions Content 1. Challenge 2. Systemic view 3. Model based approach 4. Methodology & Tool 4. Applications (Examples) 5. Conclusions Ref. M. Janovjak, 15th April

40 durch einen modellbasierenden Ansatz, 5. Conclusions Benefit of deployment of model based approach with System Dynamics Vensim - enables to master complex manufacturing SC processes and to harvest even very high hanging fruits not achievable with other approaches. - leverages user to be in position to elaborate and select optimized solution and making sustainable decisions with minimized risk and maximized benefit. -. results in paradigm shift to facilitate pro active management of complex manufacturing SC processes. Over all, the universal applicability of the applied methodology & tool opens up to build a platform for a flexible and effective governance to master future challenges. Ref. M. Janovjak, 15th April

41 durch einen modellbasierenden Ansatz Thank You! Matej Janovjak, M.Sc. S.F.I.T. (Dipl.Ing.ETH) InSysteA GmbH, Innovative Systemic Applications, CEO University of applied Science, Basel, Lecturer for System Dynamics Tel. Mobile Tel/Fax Address Dorfstr. 56, 8212 Nohl, Switzerland Web Ref. M. Janovjak, 15th April

42 durch einen modellbasierenden Ansatz Backup slides Ref. M. Janovjak, 15th April

43 durch einen modellbasierenden Ansatz, 3. Model based approach Correctness and validity of the model - Structural fit (structure of the causal relations between variables and sub-processes) - Behavior correctness (the behavior of the real system is appropriately reproduced by the model, i.e. dynamic and logical behave, equation & operating point) - Plausibility and consistency of results (qualitative and quantitative, extreme conditions) - Validity for application (it the model is appropriate for the domain in which it is to be used and conforms with purpose, i.e. based on comparison of the experimental values and simulation results, the model enables required changes and must provide requested answers, i.e. problem solving results ) Ref. M. Janovjak, 15th April

44 durch einen modellbasierenden Ansatz, 4. Methodology & Tool Methodology capabilities and tool characteristics (in extracts) The methodological capability and the tool characteristics determine the power of the context of the modelling & simulation The methodological capability - modelling of the dynamic processes - modelling of the complex feedback systems - application field universality - suitable for study and practical management - practical to knowledge mgmt. The tool characteristics - wide range of mathematic functions - easy to model and validate - high calculating capacity - high capability to provide multivariate analysis and simulation - easy customisable front-end cockpit - powerful and easy customisable graphical presentation - integrated data management (data collection, bi-directional IM/IT integration, ) Ref. M. Janovjak, 15th April