Process Optimisation - Bullant Filters
Bullant Filters are used on flow lines where a bottleneck (The Drum) is resident within a group of many interconnected and automated workstations, as shown below.
Bullant Filters rely on two Theory of Constraints (TOC) concepts;
to help characterise the types of flow lines generally found in the processing industry.
We enlist the help of operations personnel and use quantitative and qualitative statistical methods to gather powerful data and create a process map using the three Bullant Filters, which include;
Like the alignment of holes in a pile of swiss cheese slices, these filters can then be used to identify those Performance-Centred Improvement opportunities that add the maximum value.
The Bullant Filters can be customised to the clients’ needs and can be used to clarify a single, mission critical reliability issue right through to a longer term, supply chain performance improvement strategy. Please check out our white paper if you want to learn more about Bullant Filters.
We call the constraint or bottleneck of a flow line, the Drum. The Drum of the flow line shown below is Machine #3, it is the machine that determines the drumbeat (the instantaneous rate portential) for the schedule.
The capacity of the Drum relative to the other machines in a line determines the performance characteristics of that flow line.
For example the flow line below is “unbalanced” meaning that the machines upstream and downstream of the Drum (ie. Machine #3) have greater capacity. This means that they will be less likely to starve or block material to or from Machine #3. Therefore the efficiency of the line will be relatively high when compared to a line where all the machine capacities are the same as Machine #3 (i.e. “balanced”).
Ropes & Control Points
Rope is a term used to describe the gating mechanism used throughout a supply chain. It is a supply schedule at the supply chain level, it is the factory schedule at the factory level and it is the updated daily sequence at the line level. Complex supply chains need a gating mechanism to ensure that inventory and time buffers do not get out of control.
Information at Control Points are used to supplement information on buffer penetrations.
For example, in the line below, Control Point #2 can be used to monitor the upstream buffers before Machine #3. If there is no material at Control Point #2 for more than say 5% of the time then this would indicate that there is a problem with the upstream process.
The source of the problem can be pinpointed by looking at the upstream buffers (e.g. which ones are in green for a lot of the time) and control points (e.g. which control point shows 0% time without material). This information would then help to establish the upstream performance bottleneck.
The metrics used to monitor the performance of a flow line are summarised in a document that we call the Demand Filter.
A Demand Filter includes the following metrics;
- Line Effectiveness (similar to Lean’s Overall Equipment Effectiveness (OEE))
- Drumbeat (instantaneous rate at the bottleneck – The Drum)
- Running Efficiency
- Financial performance measures including Octane.
- Product Cycle Rates
The Demand Filter is used to determine the most important products as measured by the time required at the Drum of a production line.
Bullant Demand Filters have evolved as a result of many applications in different process industries. They are also used as foundation data for Value Stream Maps.
Determining the demand concentration at the Drum of a flow line using a Demand Filter is important however, the line must be designed and improved so that capacity bottlenecks do not constrain the demand.
We use a capacity filter to document the capacity distribution of machines and buffers used throughout a flow line.
A Capacity Filter includes;
- Demand Filter summary information at the Drum
- Distribution of Capacity and Time Buffers for different product configurations
- Buffer capability (from Dynamic Modelling)
- Yield loss and giveaway
- Labour concentration
The Reliability Filter is used to establish the running efficiencies for each possible running configuration of a line. It is based on information from the Capacity Filter.
Running efficiencies defined by the the Reliability Filter are then fed back into the Demand Filter so that total line running times can be confirmed.
There are 4 techniques that can be deployed as part of the Reliability Filter;
We use dynamic modelling to demonstrate how the performance of the Drum on a flow line is effected by the performance of upstream and downstream equipment and buffers.
Historical downtime data is used to understand current performance of different workstations in a flow line.
We use the term stability as well as reliability to recognise that there are often points of local choas in high-change, high technology processing lines. This is where (what Edwards Deming called) “special-cause or attributable-cause variation” remains hidden, immune to broad-based continuous improvement initiatives that focus on “common-cause variation”.
The Reliability Filter uses a Criticality Assessment tool to establish the points of instability in a line. Points of instability are those parts of a production line where repeatable operation cannot be guaranteed and where normal reliability techniques are almost impossible to apply.
A stability map is used to visual represent the points of instability on a flow line.
Broad-based continuous improvement activities in high-technology factories, while crucial, often consume lots of resources and struggle to exceed inflation in terms of budget reduction. Improvement programs are themselves becoming an efficiency issue as resources are squeezed and complexity increases.
We have coined the term Performance-Centred Improvement to help distinguish them from continuous improvements.
Performance-Centred Improvements are opportunities that appear as constraints in all three of the Bullant Filters. They are high-impact and can shift the performance of a complex line in a substantial way while complementing the sustaining effects of a broader continuous improvement effort.