Using Ashcroft Tables in Machine Assignment

Using Ashcroft Tables in Machine Assignment

• Cronometras Team

A practical guide to Ashcroft Tables for machine assignment: service ratio, binomial model, idle time and machine stops, implementation steps, sector use cases and limitations.

Pillar 1: Theoretical and Scientific Foundations

The Classic Multiple Machine Assignment Problem

Methods and time engineering is constantly faced with a fundamental efficiency dilemma: the assignment of human resources to semi-automated processes. This challenge, known in the technical literature as the Machine Assignment Problem, is the core of the Ashcroft Tables. Their correct application remains a pillar for optimizing technical productivity in discrete manufacturing environments.

Definition of the Machine Assignment Problem in methods engineering

Essentially, the problem arises when an operator must supervise or intervene on several machines that alternate automatic cycles with manual operations. The central question it seeks to answer is quantitative: how many machines should a single operator attend to achieve the best balance between labor utilization and equipment utilization?

This is not a mere matter of visual supervision. It involves active service times (loading, unloading, setting, cleaning) that compete for the operator's attention. A suboptimal assignment generates two types of critical waste in Lean: Muri (overburden) and Muda (waste of idle time or machine capacity).

Dual objectives: minimize operator idle time and machine stoppages

The model seeks to optimize two conflicting metrics:

  • Operator idle time: The percentage of time the worker is unoccupied, waiting for the machines to complete their automatic cycles. A high value indicates underutilization of labor.
  • Machine stoppages: The percentage of time the equipment remains inactive, waiting for the operator's manual intervention. A high value directly impacts productive capacity and cost per piece.

The perfect balance minimizes the sum of both idle times, maximizing the overall efficiency of the man-machine system.

Relevance in OEE (Overall Equipment Effectiveness) optimization

OEE is the gold standard for measuring equipment productivity. It consists of Availability, Performance and Quality. The Ashcroft Tables have a direct impact on the first pillar: Availability.

A machine stoppage due to lack of an operator is an unplanned availability event. By using the tables to assign the optimal number of machines, the probability of these stoppages is statistically reduced, increasing the availability factor of the OEE. Therefore, Ashcroft is not an isolated tool, but a fundamental component of the plant's measurement and continuous improvement ecosystem.

Probabilistic Basis and the Ashcroft Binomial Model

The elegance of the Ashcroft model lies in its statistical simplicity. It transforms a complex operational problem into a manageable binomial probability calculation, based on two clearly defined input variables.

The two key variables: service ratio (r) and number of machines (n)

The entire model pivots on these two parameters:

  • Service ratio (r): The proportion of the machine's total cycle that requires active attention from the operator. It is calculated as r = t_s / t_c, where t_s is the manual service time per cycle and t_c is the total cycle time (automatic + manual). A low r indicates highly automated machines.
  • Number of assigned machines (n): The variable we want to optimize.

The relationship between r and n is inverse. The lower the service ratio (the more automated the machines), the more machines a single operator can attend without creating bottlenecks.

Fundamental formulas and probability calculation

The model assumes that each machine, at any random instant, has a probability p of being in its automatic phase (it does not need the operator) and a probability q = 1-p of requiring manual service. Here, q is equivalent to the service ratio r.

The probability that exactly k machines out of the n assigned require simultaneous service follows the binomial distribution: P(X = k) = C(n, k) * p^(n-k) * q^k

The tables precalculate the expected values of idle time and stoppages for different (r, n) pairs, sparing the engineer the tedious summation of probabilities.

Interpretation of the output parameters: idle time and stoppages

The table results are not just raw percentages. Their interpretation requires engineering judgment:

  • % Operator idle time: Indicates the slack in the labor assignment. A value of 5-8% may be acceptable, considering unforeseen events and breaks. A 15% figure suggests that another machine could be assigned.
  • % Machines stopped: This is the critical capacity metric. In high value-added processes (e.g., injection molding of technical plastics), even 1% of unplanned downtime can justify reassigning resources.

The engineer must compare these values with the organization's thresholds and with industry standards.

Structure and Reading of the Ashcroft Tables

The classic Ashcroft tables are double-entry matrices. Mastering how to read them is essential for their correct application on the shop floor.

How to navigate the combinations of service ratio and assigned machines

The typical structure presents:

  • Rows: Increasing values of the service ratio r (e.g. 0.10, 0.15, 0.20, ..., 0.50).
  • Columns: Number of assigned machines n (e.g. from 2 to 10).
  • Cells: Contain two main values: the % of operator idle time and the % of machine downtime.

To use the table, the actual r of the operation is first calculated through a stopwatch time study, a tool that platforms such as Cronometras have digitized and greatly simplified. Then, the corresponding row is located and read horizontally to see the effect of assigning 2, 3, 4... machines.

Practical application example with r=0.30 and n=4 machines

Suppose a deburring operation in a metalworking cell. The time study yields a t_c of 1.50 min and a t_s of 0.45 min. The ratio is r = 0.45 / 1.50 = 0.30.

Looking up the table for r = 0.30 and n = 4:

  • Operator idle time: ≈ 6.0%
  • Machine stoppage time: ≈ 1.3%

These values are considered highly efficient. The operator has slight slack for unforeseen events, and the machines almost never wait. Assigning only 3 machines would significantly increase the operator's idle time (labor inefficiency). Assigning 5 machines, on the other hand, would sharply increase the machine stoppage percentage, reducing productive capacity.

Typical system efficiency thresholds

In industrial consulting practice, the following reference ranges are commonly used:

  • Optimal system: Both parameters below 5-7%.
  • Acceptable system: Operator idle time between 7-12%, machine stoppages below 5%.
  • System to review: Any parameter consistently exceeds 10-12%, indicating an imbalance that requires reassignment or method improvement.

These thresholds are flexible and depend on the sector, the criticality of the process, and labor and equipment costs.

Pillar 2: Practical Application in Industrial Environments

Step-by-Step Implementation in an Industrial Plant

Theory finds its value in methodical application. Follow this protocol to implement the Ashcroft analysis in your production area.

Data collection: cycle times (tc) and manual service (ts)

This is the most critical step. The input data must be representative and accurate.

  1. Select the operation: Identify a workstation where one operator attends several identical machines or machines with similar cycles.
  2. Conduct a time study: Use a stopwatch or a digital application to measure, over at least 20-30 cycles:
    • t_c: Total time from the start of one cycle to the start of the next.
    • t_s: Net manual intervention time (load part, start cycle, remove part, inspect). Work sampling, facilitated by applications such as WorkSamp, can be a useful complement to validate the proportion of service time relative to other work elements.
  3. Calculate the ratio r: Average the times and apply the formula r = t_s_promedio / t_c_promedio.

Calculating the service ratio and consulting the tables

With the value of r, the Ashcroft table is consulted. A leveling factor (normally between 1.05 and 1.15) should be considered to absorb the normal variability of the process. If r = 0.30, we could use r_ajustado = 0.33 to be conservative.

The corresponding row is explored for different values of n. The objective is to find the n that offers the best compromise according to the plant's thresholds.

Validation through observation and time study

The theoretical result must be validated in the field. The proposed assignment (e.g. 4 machines per operator) is implemented during a pilot period (1-2 shifts). During this period, the following are timed and recorded:

  • Actual machine stoppages due to lack of an operator.
  • The operator's observed periods of inactivity. These observed data are compared with those predicted by the table. If there is a significant deviation (>20%), the quality of the initial data should be reviewed, or non-modeled factors considered (such as long travel distances).

Use Cases by Industrial Sector

The versatility of the model allows it to be adapted to multiple industries, provided there is a cyclical interruption of an automatic process by a manual task.

Textiles and apparel: assignment in sewing and knitting machines

In a knitted fabric plant, one operator can attend multiple circular knitting machines. The automatic cycle is the knitting of the garment; the manual service is changing the yarn bobbin or clearing a snag. With a typically low r (0.15-0.25), the tables can justify a single operator supervising 6 or 8 machines, radically increasing productivity per employee.

Metalworking: CNC lathe and press operation

In machining, the automatic cycle (turning, milling) is usually long. The manual service (loading/unloading the part, control measurement) is relatively brief. Here, r may be around 0.20-0.35. Ashcroft helps design work cells where one operator feeds 2 or 3 CNC lathes, keeping the spindles cutting almost continuously. For real-time production control of these cells, platforms such as Induly offer visibility that perfectly complements the initial time study.

Plastics injection molding and SMT electronics

In injection molding, the cycle is fixed and automatic. The service is the removal of the part and cleaning of the mold. An r of 0.10 is common, allowing one operator to attend 10 or more simple machines. In SMT lines, the operator supervises the feeding of components into the feeders of the pick-and-place machines. The model helps to size how many lines a technician can monitor.

Integration with Continuous Improvement Systems

Ashcroft is not an isolated exercise. It integrates naturally into modern production management systems.

Use in work cells and Lean manufacturing

In a U-shaped cell, where the operator can move easily between machines from different processes (lathe, milling machine, drill press), the Ashcroft model is used to justify multi-functionality. It helps answer: can this operator, with the reduced transport distance of the cell, effectively attend three machines with different cycles? The tables, adjusted with a factor for minimal travel, provide the quantitative answer.

Feeding the availability calculation in OEE

As mentioned, the stoppages calculated by Ashcroft are a direct input for the availability factor of the OEE. If the model predicts 2% stoppages due to assignment, that is the ceiling of achievable availability under that configuration. Any subsequent improvement must target other losses (breakdowns, changeovers), or reconsider the machine assignment.

Application in production leveling (Heijunka)

When production volumes and mix are leveled in a cell, the operator's workload varies. Ashcroft can be used to evaluate different machine assignment scenarios based on planned demand. It allows flexible sizing of the number of operators required for different production levels, maintaining the efficiency of the man-machine system.

Pillar 3: Current Context, Limitations and Evolution

Recognized Limitations of the Pure Model

Like any simplified model, Ashcroft has assumptions that are not always met in complex industrial reality. Recognizing them is the first step to using it with judgment.

Assumption of independence between machines

The model assumes that a machine's need for service is independent of the others. This is valid for identical, isolated machines. However, in a sequential line (where the stoppage of one machine blocks the previous one or starves the next), this assumption fails. A stoppage can have a domino effect that the binomial model does not capture.

Real variability vs. the model's deterministic times

Ashcroft works with deterministic average times. In practice, both t_c (due to variability in the automatic process) and t_s (due to the varying complexity of the manual task) follow probability distributions (lognormal, Weibull). This real variability, especially when high (coefficient of variation > 20%), can cause actual results to deviate significantly from those predicted by the table.

Factors not contemplated: travel and prioritization

Two important omissions:

  1. Travel times: The model assumes the operator can attend to any machine instantly. In reality, the walking time between machines can be significant, especially if they are far apart. This time must be subtracted from the time available for service, increasing the effective r.
  2. Machine prioritization: Not all machines are equal. A high-tonnage stamping press may have a much higher cost per minute of downtime than a drill press. The model does not differentiate; it treats all stoppages equally. In practice, the operator will prioritize the most critical machine, altering the probabilities.

Evolution Toward Hybrid Models and Simulation

Modern engineering does not discard Ashcroft, but empowers it by integrating it into broader and more sophisticated frameworks.

Complementing with discrete event simulation

Simulation is the perfect tool to overcome the limitations of Ashcroft. A model can be built in software such as Arena or Plant Simulation where:

  • Cycle and service times follow specific probability distributions.
  • The operator's travel between machines is modeled.
  • Prioritization rules are implemented.
  • The system is simulated over thousands of virtual cycles.

The Ashcroft results serve as an excellent starting point and as validation of the simulation model. If both agree, there is high confidence. If they differ, the simulation reveals the factors (variability, distances) that cause the discrepancy.

Reformulation as queueing systems

The multiple assignment problem can be viewed as a queueing system where:

  • Customers: The machines requiring service.
  • Server: The operator.
  • Queue discipline: Can be FIFO (first to arrive, first to be served) or prioritized.

Under this framework, M/G/1 (single server) or M/G/c (several operators) queueing theory models are applied. This makes it possible to calculate metrics such as the average waiting time of a machine to be attended, something the Ashcroft binomial model does not provide directly. Integrating sampling tools such as WorkSamp to characterize arrival and service times (the "G" distributions) is key to feeding these queueing models with real data.

Conclusion

The Ashcroft Tables remain, more than 70 years later, a top-tier tool in the methods engineer's arsenal. Their strength lies in providing a quick, objective quantitative basis for an everyday problem. However, the modern engineer must use them with discernment, aware of their assumptions and limitations.

The future of machine assignment lies in hybrid models: the analytical elegance of Ashcroft for initial sizing, combined with the power of discrete event simulation for validation and fine optimization in complex scenarios. This approach, anchored in accurate time study data, is what makes it possible to reach new levels of excellence in technical productivity and OEE.


Resources and Tools

To delve deeper into the practical application of these concepts, we recommend the following resources:

  • ASETEMYT - Industrial Time Study Directory: Find tools, software and services specialized in work measurement.
  • Cronometras: Specialized consulting in methods and time engineering, productivity system implementation.
  • Induly: Production Control and Industrial Time & Attendance software to obtain real-time data from your plant.
  • WorkSamp: Application for conducting agile and accurate Work Sampling studies.
  • ASETEMYT Blog: Technical articles on time study, OEE, MTM and process improvement.
  • Add your company to the directory: If you offer services or products in the field of time study and industrial productivity.

Don't want to do it yourself? We do it for you from €800

If you prefer to outsource it, Cronometras delivers a complete time and motion study from a single video: elemental breakdown, OIT/TAL fatigue allowances, standard time with tolerances, and a report in PDF + Excel + JSON ready to present to management.

👉 Request the study from €800 · More about Cronometras

Did you find this useful?

Share this article with your professional network