Design & Improvement of Manufacturing Systems through the use of Factory Physics:

1.Introduction:

-> DEF: FACTORY PHYSICS is a systematic description of the underlying behavour manufacturing systems. It…

• … identify opportunities for improving existing systes;
• … design effective new systems;
• … make the tradeoffs needed to coordinate policies from disparate areas.

Source: Wallace, J.H., Spearman, L.M.. “Factory physics: physics: foundations of manufacturing management”. Irwin McGraw-Hill, 2001.

-> OPERATIONS MANAGEMENT (OM): is the tradisional field in which operations are studied.

• Is a technical perspective compared to other operations management in other areas.
• Operations as applications of resources to produce good services:
• Operations as focus on the flow of materials through a plant.

-> CHAR:

• Focus on manufacturing management decisions:

• Provide a perspective on manufacturing operations as:
  • Applications of  resources to the production of Physical goods;
  • Flow of materials through the plant.
• Each manufacturing environment is unique:
  • No single set of work procedures can work well under all conditions;
  • Manufacturing managers have to rely on a general understanding of their manufacturing systems to enable themo to identify leverage points, creatively overcome the competition & engender an environment of continual improvement.

-> BACKGROUND LAWS:

• Behavioral tendencies shared by virtually all manufactured systems (operations POV);

-> Can be organized into a body of knowledge to serve as a manufacturing manager’s knowledge base.

• Establish basic concepts as building blocks, starting fundamental principles as “manufacturing laws” & identify general insights from the specific practices.

Basic Laws of Production Logistics.

-> DEF: universally applicable correlations between production logistics objectives & correspondingly, between key performances.

• Can be applied to assess qualitatively and quantitatively the impact of individual behaviour with regards to the logistic objectives of the production system.

-> The problem solving is built upon the analysis of the behavior of the manufacturing systems (factory dynamics) based on the used of the “laws” of factory physics.

2.Terminology & Scope of Work:

Some Terms…

• Workstation (or station or work-center or …): a collection of one or more machines, or a collection of one or more manual stations, that perform (essentially) identical functions.
• Part: a piece of component, sub-assembly, or an assembly that is worked by, and moves through, the workstations.
• End item (or finished product): a part that is sold directly to a customer (remark: relationships between end items and their constituent parts is maintained in the Bill of Material).
• Routing: a sequence of workstations passed through by a part (i.e., paths through the plant).
• Order (or customer order or replenishment order or …): a request for a particular part number, in a particular quantity, to be delivered on a particular date, being it either an actual or forecasted customer order, or a replenishment order.
• Job (or work order, or production order): a set of physical materials that is transferred through the workstations (i.e. that traverses a routing), along with associated logical information.
• Throughput or throughput rate (TH): the average output of a production process (machine, workstation, line, plant) per unit time (e.g., parts/minute, jobs/hour).

-> Other uses:

• At firm level is defined as the production per unit time that is sold.
• For a plant/ workstation is the average quantityof good (nondefective) parts producewds per unit time.

-> Lines with tandem WS dedicated to a single family of products, where all products pass through each station exactly one, the TH of very station will be the same (without loss);

-> In plant where workstations service multiple routings, the TH of an individual station is due to the mix of routings/ product types passing through it.

• Capacity (or maximum capacity): the upper limit on the throughput of a production process.

-> Most cases, releasing work into the system at or above the capacity causes the system to become unsatable.

• Work In Process (WIP): the inventory between start and end stock points of a product routing.
• Throughput time or flow time or sojourn time (TTP): average time span a job requires from its release at the beginning of the routing to the end of the routing, when it reaches the end stock point (note that TTP can be measured also for single parts flowing through a system).
• Delivery lead time: the time allotted for production of a job on that routing.

Job-Shops:

-> CAHR:

• small batches/lots are produced with a higher variety of routings through the plant;
• production flows through the plant are jumbled;
• jobs are characterized by specific (even unique) processing requirements;
• setups are commonly occurring to prepare the workstations for new jobs.

Connected lines:

 -> CHAR:

• product routings between workstations are fixed ones;
• workstations are connected by a paced material handling system (i.e. intermittent and continuous flow lines);
• parts/products are processed and moved as a single piece or even in (very) small transfer lots (within single-, multi-, mixed-model production management).

Disconnected flow lines

-> CHAR:

• parts/products are produced on a (limited) number of identifiable routings;
• product routings can be fixed ones but also even (partially) distinct ones;
• workstations are not connected by a paced material handling system;
• inventories can build up between workstations;
• batch production flow (i.e. the entire batch/lot is ended before moving to the next workstation) or single piece flow (i.e. a single part is moved at a time to the next workstation) or intermediate cases (as a small transfer lot, portion of the entire batch/lot) are all possible cases.

The Funnel Model – Little’s Law:

-> HP:

• Steady system.

-> LAW:

TH = WIP / LT

-> Where:

• TH = Throughput: average output of the process per unit of time;
• WIP = Work in progress: inventory processed or waiting in production;
• LT = Lead time (or throughput-time or flow-time): average time between release of an order into the production system and its completion (LT is the time period inter-current between the input time of a part into the system and its output time from the system itself);
• WIP* – critical WIP: the WIP level at max. throughput Thmax.

-> CHAR:

• Allow to understand relationship among WIP, Throughputm TH & LT of parts flowing in a generic system for discrete production, give the input rate of clients is constant.
• Is independent of the configuration of the system, type of distribution of processing times, of routing & of the distribution of inter-arrival times.
• 🔺 Input Rate of parts => 🔺 WIP, 🔺 TH, Lt is constant.

-> Critical WIP is reached => TH achieve its maximum & never increase more, LT start increasing;

-> If 🔻 LT + TH costant => we have to 🔻 WIP.

Line Description:

BOTTLE NECK (): Rate (parts/unit time or jobs/unit time) of the station having the highest long-term utilization.

• We have to understand if we are performing well.

RAW PROCESS TIME (): Sum of the long-term average process times (working time) of each station in the line.

CONGESTION: different cases may happen:

• Best-case performance (zero variability, zero randomness): we do everything we have to do as soon as possible;
• Practical worst-case performance (maximum randomness): all pieces are processed between the two possibilities;
• Worst-case performance (zero randomness but batch moves): we do everything we have to do at last time;

⚠️Lines with same  &  can have differen behave.

Relationship > CRITICAL WIP ():

-> DEF: the WIP level in which a line having no congestion would achieve maximum throughput (i.e., ) with minimum throughput time (i.e., ).

The Penny Feb:

-> Features

• 4 identical machines in line (4 stations)
• Process time of each station 2 h
• Process time are fixed (no variability = steady state
• Inter-arrival time of parts  (Delta T) is lowered step by step
• Elapsed time (Time) is registered

Best Case Law:

-> The minimum throughput time (i.e., TTPbest) for a given WIP level, w, is given by:

-> The maximum throughput (i.e., THbest) for a given WIP level, w, is given by:

📌Little’s law valid over the long term.

Worst Case Law:

-> The worst case throughput time (i.e., TTPworst) for a given WIP level, w, is given by:

-> The worst case throughput (i.e., THworst) for a given WIP level, w, is given by:

Practical Worst Case (PWC) Law:

-> Let w = jobs in the line, N = number of stations in the line, and t = working time at each station:

-> Using the Little’s law       TH = WIP/TTP      =    [w/(W0+w-1)] * rb

• W0: is the critical WIP;

📌Dimostration not needed.

-> The practical worst case throughput time (i.e., TTPPWC) for a given WIP level, w, is given by:

-> The practical worst case throughput (i.e., THPWC) for a given WIP level, w, is given by:

-> 🔼# parts of system <=> 🔼 TP

Penny Fab:

->TH and TTP of the PWC are placed between those of the best case and of the worst case; as such, the PWC can be interpreted as a midpoint usable to approximately judge the behavior of the real system. Indeed, an “internal” benchmarking methodology can exploit the three cases to this end.

-> By collecting data on the average WIP, TH, and TTP of the real system, it can be detected whether its operating conditions lie in the region between the best and the practical worst cases, or between the practical worst and the worst cases.

TH vs. WIP : Best-, practical worst- and worst-case performance

TTP vs. WIP : Best-, practical worst- and worst-case performance

Conclusion:

-> Concluding Factory Physics suggest:

• To describe the underlying behavior of manufacturing systems in particular operating conditions (i.e. the best, the worst, the practical worst case);
• To model or to monitor the real production system in given (i.e. expected or actual) operating conditions, in order to measure the real system performances and their correlations;
• To decide based on the performance assessment of the system in its expected / actual conditions, and the “internal” benchmark provided by the best, worst, practical worst case.

System Variability:

-> DEF VARIABILITY: anything that causes production system to depart from a regular & predictable behavior.

• Subsequent to a decision or to an unpredictable randomness.

-> SOURCES of VARIABILITY:

• Natural, includes minor fluctuations in the working time due to differences in operators, machines & materials;
• Random & Preemptive outages, include machine failuress, operator unavailability, shortages of consumable;
• Non Preemptive outages, includes setups, preventive maintenance, operator meetings, breaks, reworks.

-> Relationship to performance cases:

• LV: Between Best-case & Practical worst-case;
• MV: Practical worst-case;
• HV: Between Practical worst-case & Worst-case.

Workstation Performances:

Production Flow:

LOW UTILIZATION => propagation of flow variability don’t affect the flow;

HIGH UTILIZATION => propagation of flow variability affect the flow;

Work Content Model:

-> DEF: is a measure of the working time…

• Required by an operatoin on a workstation;
• Working time or process time are used as synonymous.

-> Can:

• Used as measure of the planned time;
• Used as feature describing the production mix;
• Defined by using sample data over a reference period.

-> Work Content of a job is measure of the working time

*Required by an operator on a workstation;

**Working time or process time are used as synonymous.

Parameters:

Performance Measures:

-> For each WS an Operation Time (TOP) can be calculated:

OPERATION TIME:

-> DEF: can be calculated: it expresses the time needed in order to complete the job (production order, campaign, …), measured in the shop calendar.

• It’s defined for each planned operation in the production process, similarly to the work content WC.

-> ROUT express the maximum working hours per day (Shop Calendar Day), provided by a workstation (or each single machine or manual station, in the workstation) to perform the requested jobs.

• It is a “maximum” rate, defined assuming the maximum efficiency, without performance losses.
• Thus, TOP expresses how many SCDs a workstation / a machine / a manual station is busy, under the hypothesis that it works at its maximum capacity.

-> Each workstation is characterized by a throughput that is nominally achievable; the nominal value of the throughput is, under the hypothesis of maximum efficiency, equal to (for the i-th workstation):

-> Station with higher Utilizatoin Rate is the bottleneck.

-> The maximum capacity of the production system is constrained by the bottleneck (i.e., the workstation with maximum TOP, alias minimum TH).

-> Each workstation, made of machines (manual stations) characterized by different throughput (j-th different types within the i-th workstation):

-> The maximum capacity of the production system is still constrained by the bottleneck: