Assembly line balancing can be thought of as a process for optimizing assembly lines to address certain factors, such as improving quality control. Organizing an assembly line can be a complex problem, depending on the industry of manufacture, but optimizing any system for better results is essential to all business models. Operating and maintaining assembly lines can involve considerable expenditure of time and money through labor and equipment. The goal of balancing them is to maximize operations in whatever areas need improvement, all of which ultimately improve margins.
As one example, a company that builds power boats is facing higher demands and might want to focus priority on speeding up assembly line production. They would start by investigating the assembly line layout. There are a given number of work stations to complete the process, and the steps taken at each work station require a certain amount of time and labor to complete. Before expanding assembly lines and hiring more workers, the company will look at ways to reduce non-essential time or combine steps.
Assembly Line Balancing is about the leveling the workload across all processes in a cell or value stream to remove bottlenecks and excess capacity in an assembly line.
The result of effective assembly line balancing will be improved efficiency and an increase in units produced over a given time frame without sacrificing quality. The different steps involved in the manufacturing process might involve different skills, raw goods, and tools, so the company works within certain restrictions.
Optimizing these specific steps will require different approaches, including recalculations of distance traveled, steps followed, available technologies, and so on. More complex manufacturing can be broken into essential segments to be examined separately. Individual times and use of resources may be analyzed in especially complex manufacturing environments such as those demanding customization or high-precision assembly. Balancing guides business decisions on altering the variables that may be involved. Final decisions might be based on comparisons of several different mathematical approaches to weigh their value in terms of improved efficiency vs increased risk of error or lower quality.
A Balanced Model
The general principle for approaching assembly line balancing is based on the number of fundamental tasks involved and reducing the demands of each task. The concept should be a linear process without overlap or recursion. Production time and cycle rate might be given as ideals or as averages for certain tasks, including quality checks, but generally do not involve splitting tasks. Heuristic rules for creating balanced assembly lines fall into five areas for ranking process stations, from lowest to highest priority:
- 1. Process or quality importance
- 2. Shortest time of operation
- 3. Most following tasks required
- 4. Longest time of operation
- 5. Least number of following tasks required
Balancing is a more flexible approach on this ranking. For example, if the longest time of operation won't fit into the allotted work flow time for a particular station, try the next longest time of operation. In many cases, the order of operations will have to follow a fixed sequence, such as steps for preparing the chassis in an automobile production plant as it will be the foundation for other parts and it might be counter-productive to introduce it later in the process.
If two tasks have similar times and priorities, ranking them is purely arbitrary and should fall back to their position in the assembly sequence. Ranking stations by their complexity, importance, and time frames will help to prioritize areas having the greatest impact, and thus providing greater returns through improvement.
Cycle Time: The cycle time for the completion of the process can be taken from actual times of execution or seen as a proportion of the demand rate to maintain workflow.
Unit of Time: For better clarity and ease of calculation, the unit of time employed should be the same for all steps: seconds, minutes, or hours.
Needed Time: The time to complete each task might be the minimum or average time recorded at each station, taking into account how consistent controls are.
Allocated Time: The time allocated to each step to maintain work flow.
Precedence: Any preliminary tasks and times that must be taken before a particular operation can take place, including quality checks.
Idle Time: Any downtime that takes place between completion of the necessary steps and progression to the next station.
Ready Task: This is any step that has met all precedence's but has not yet been picked up in the work flow in sequence.
These help to determine the three critical results:
Efficiency: This is the time allocated divided by the time needed. For instance, a window of 15 seconds allocated for each task completed in 17 seconds could be given as an efficiency of 88 percent. Where the values are reversed, say 20 seconds of allocated time vs 18 seconds of needed time, you're looking at an efficiency of 111%, which is not realistic. The conclusion to be drawn is that 2 seconds are being wasted in this process and less time should be allocated. If this 18 seconds is an anomaly and not a mean time, more streamlined processes are needed to avoid such variation.
Balance Delay: This represents the percentage of total wasted time, to be measured for all work stations and for total unit production over the course of the assembly process.
Theoretical Work Stations: This is a count of necessary work stations arrived at by dividing total unit production time by each task time. If production of each unit takes 5 minutes, and your longest station time needed is 1 minute, you should need no more than 5 stations maximum for the assembly process.
Computing Production Time
A crucial step involves diagramming the entire process in terms of sequential stations, with the time required for each and the total time of production. This should produce a formula for estimating production times in terms of volume. For example, a demand of 2,000 units scheduled for 8 hours of production time (2,880 seconds) would require producing one unit every 1.44 seconds. How achievable that is depends on the processes involved, but it's easy to demonstrate whether or not it's a realistic expectation based on balancing.
Problems arise when the time needed to perform a certain task is greater than the time available. This may require splitting tasks. This amounts to duplication of work. This may seem less efficient, and may require additional investment, but is often the only way to make certain stations align with time frames. For example, a station which requires 6 minutes to perform a task within the 2-minute window of the next operation would be split into 3 stations. While three different employees at different levels of skill and experience may involve some variation, the ability to produce 3 units every 6 minutes will meet the demand for another unit every 2 minutes to eliminate idle time at the next station.
These assumptions may vary from process to process. For instance, time and expense is involved for every employee "touch" of the product as well as the distance required for the product to be moved. Coordinating 3 separate stations to deliver 3 separate products may represent inefficiency where tolerances of time and cost are very tight. The expense of hiring personnel and buying or leasing equipment for two additional stations may not be worth saving a few seconds or a few pennies. However, even such a slim gain might be acceptable in long-run, high-volume situations where positive returns will be seen much sooner.
One of the main objectives to keep in mind for assembly line balancing in production is the integration of quality control measures. Quality standards will vary between companies and industries. Some manufacturing processes will involve precise measurements and testing at various phases of assembly to ensure performance or safety. Other businesses might approach quality as a brand necessity. But where quality checks are essential, a unit may not pass to the next station without QC approval.
Quality tests at different stations should be integrated into the allocated time for each station, or delegated to an independent station in terms of time frames. It depends on how long that testing process (and documentation) takes. Quality checks taking 2 minutes may be an unacceptable delay on production phases where actual assembly takes a similar or even shorter interval. In those cases it makes sense to maximize QC effectiveness without impeding production, typically by inspecting 1 out of every 10 items, on a materials batch basis, or at different points in time, such as the first few items in a run and subsequently intermittent or random testing. Other quality checks might involve disassembly or more exhaustive checks of finished items that don't interfere with work flow.
Of course, the more items that are not inspected the greater the risk of passing on product that does not meet standards. Assembly line balancing in production requires establishing minimum quality standards first and implementing the procedures that will best satisfy them with the least impact on production.