We are pleased to bring you this article by Cornelio Abellanas, a practitioner of lean manufacturing who is based in Spain. His strong technical background validates what many practitioners of lean manufacturing take for granted, but he knows through mathematical validation how and why certain lean methods work.
Enjoy this article and learn more about Cornelio at the end of this article.
The purpose of Lean is the elimination of waste while Six Sigma is centered in the reduction of variability. One of the forms of waste we try to eliminate with Lean is excess inventory and particularly Work-In-Process which is responsible for long lead times and poor customer responsiveness.
Variability was described by Myron Trybus as a virus which infects our processes: it causes chaos and is often undetected. Variability is in fact often the root cause of our problems and Six Sigma allows us to detect it and reduce it at the source.
The best way to understand the effects of variability is by using a Monte Carlo simulator. In my teaching I use a simple 3 step process simulator in Excel (shown above) to let participants experience for themselves the effects of variation, try different solutions to the problems presented and see the side effects by downloading the monte carlo simulation yourself.
We want to see the effects of variation in one step on the total process. To do that we first run the simulator for a while by pressing start and see what the ideal process would look like: 100 items committed and delivered to the customer on every period with an average WIP per step of 100 and average lead time of 1.
Now let us key in a variation of 40 in step 2 (random variation of thruput between 60 and 140 with an average of 100). When we run the first thing we experience is customer dissatisfaction due to missed deliveries (% On time drops).
Eventually some WIP starts accumulating before and after step 2 (the one causing the problem). If we continue we will see WIP moving between Before and After step 2. When the excess WIP is After it protects the customer from the variability so % On time recovers to 100% but when the excess WIP moves to Before we start missing deliveries again.
In the graph in the bottom we can see the evolution of average lead time (which is related to total WIP) as well as the instances of missed deliveries. The first thing we notice, as a direct consequence of variability is an increase of WIP which in turn causes longer lead times. The "virus" effect of variability can also be noticed when we look at the average thruput of step 2: it is still the same as that of the other steps, therefore it is not obvious where the problem is coming from. We can also see that on some occasions the excess WIP accumulates AFTER step 2 and this can mislead us to blame step 3 (it has the big pile in front of it).
This result confirms that unless we measure variability we will not improve it (Six Sigma) and the effects of variability is excess WIP and long lead time (Waste).
In an upcoming article, we will later experience the simulation will show when we apply just in time (JIT) to the model.
About Cornelio Abellanas
Has a PhD in Telecommunications Engineering degree from the University of Madrid and a Master of Science from the University of Kent at Canterbury (UK). He is a Lean Six Sigma Black Belt and EFQM (European Foundation for Quality Management) assessor.
He has been Lean Six Sigma manager and a member of the Management Committee in Celestica Spain where he implemented 80 improvement projects per year with a total savings for the Company of 1% of the yearly revenue. Improvement projects yielded these savings by reducing operator time, optimising plant layouts, reducing equipment setup time, reducing admin and production lead times, increasing equipment availability, reducing operator and equipment defect rate, reducing work-in-process and component inventory, reducing scrap and rework, improving on-time delivery, etc.
He designed and implemented a line data collection system and real-time operator feedback which enabled self control in autonomous electronic board production cells.
He performed internal audits and achieved success in external audits for the Company with standards ISO 9000: 2000, AS9100 (Aerospace), ISO/TS 16949 (Automotive) and ISO 13485 (electro-medicine). He has developed and maintained the Quality System: Quality Manual, Process Value Stream Maps, Procedures, Corrective actions, Suggestion program, etc. for Celestica Spain. He has delivered EFQM training and assessment in companies Volvo Truck and Guzman.
He currently delivers Lean and Six Sigma training and leads Kaizen workshops in companies around Europe: IBM, AT&T, MSL, Philips, Volvo Truck, Ford, Celestica, IBC, Pt Pro, BP Solar, Italgres, PCS, General Dynamics, Telefónica International Wholesale Services, EMT Valencia, Importaco, Faurecia, Asac Pharma, etc. in France, Spain, Portugal, Belgium, Netherlands, Germany, Austria, Italy, Ireland, Greece, Switzerland, Sweden, etc.
He lectures at MBAs in local Universities.
He has managed and delivered courses at the IBM International Education Center in La Hulpe, Belgium on the topics: Total Quality Management, ISO 9000, EFQM, Stress Screening, Six Sigma, Statistical Process Control and Capability, Design of Experiments, Design for Manufacturability, Business Process Improvement, Lean Production, Kaizen, Theory of Constraints, etc.
He managed an Independent Business Unit which was responsible for production of a tape drive unit designed in IBM Tucson, USA and manufactured in a local IBM Spain subcontractor. He was Manufacturing Engineering Manager for banking control units in IBM Spain. He designed a matrix ticket printer for British Railways in Ventek Ltd., London (Datapoint Computers representative for UK). He worked as a Systems Engineer in logic design for Burroughs Machines Ltd. (now Unisys) in Cumbernauld (Scotland).
Contact information:
Polyhedrika CB
Valdelinares 2, 11a
46015 VALENCIA (Spain)
Mobile: (34) 678464624
Modeling The Impact of Variability with Monte Carlo Simulation is a post from: Lean Six Sigma Consulting
search terms for this article:
queue theory in operations research, monte carlo simulation, what is a queueing theory, risk management aluminium smelter monte carlo, monte carlo process, monte carlo excel, what is monte carlo simulation, production facility model monte carlo excel, montecarlo sigma media, Monte carlo variability, monte carlo six sigma, monte carlo simulation queuing theory in operations, monte carlo simulation program, monte carlo simulation model, queueing theory and monte carlo simulation, queuing montercarlo problem, variability simulation in excel, variability simulation, variability lean manufacturing, varaibility causes in six sigma, USING QUEUING ANALYSIS AND COMPUTER SIMULATION MODELING TO REDUCE WAITING TIME, steps in monte carlo simulation in operations research, six sigma monte carlo], six sigma monte carlo, six sigma in monte carlo, simulation monte carlo excel, Simulating impact of variability, monte carlo simulation manufacturing, monte carlo simulation for manufacturing process, fun with monte carlo simulation, fun monte carlo methods, external variability in raffles, excel template monte carlo simulation, excel queue simulation, excel monte carlo simulation for queueing, effects of variability, disney fastpass waiting time reduction statistics operations research, defects effect on wip 10%, crystal ball excel models, how soon will we start to notice the effects, how to do monte carlo simulation in excel, impact of variability in manufacturing system, monte carlo simulation for excel, monte carlo simulation excel production facility model, monte carlo simulation excel production, monte carlo method 소스, monte carlo for process variability, monte carlo calculation time occasions, monte carlo analyse excel, modelling wip excel, modeling the impact of, Internal Audit - what is WIP in aluminium smelter and carbon process, aerospace component manufacturing variability dB
