The concept of bottlenecks in manufacturing has been around for at least 20 years, and many managers feel they are well educated in the need to eliminate them. It is, however, always a good idea to refresh this concept and look at it from a few different angles. The concept of a bottleneck was popularized by Eli Goldratt when he published his remarkable manufacturing book, The Goal.
Let’s start with the definition of a bottleneck:
A bottleneck occurs when the demand or need for a process exceeds the available capacity. This is characterized by the fact that time lost cannot be made up.
Image source: zubinmehta.wordpress.com
A bottleneck is not necessarily permanent, but can emerge and disappear with changing demand. For example, the welding department, that is used in the fabrication of several products for a company could become a bottleneck every time a particular product is run, but the department usually has plenty of capacity. This is why it is sometimes difficult to identify bottlenecks. Someone will think a department is a bottleneck, but the very next day the function has no demand and is idle.
It is also difficult to pinpoint bottlenecks because of the dynamic nature of any endeavor that requires lots of people, which means there can be lots of noise in the system. For example, consider the following day in the life of a particular company that makes a simple product that is made from four components fabricated in separate departments and then assembled. Each department has a different capacity. Let’s call the departments A, B, C, D and Assembly
Hour 1: Rough start - lost production in every department. Also department C has a worker that is late – lost production for the first two hours of the day in C.
Hour 2: A machine breaks down in dept B, but is fixed quickly – lost production in B.
Hour 3: Because of the breakdown in dept B in hour 2, the assembly department is idled for 20 minutes.
Hour 4: Workers are getting hungry and anxious as they wait for the lunch hour to begin. Lost production in all departments.
Hour 5: A worker is injured in dept B and is out for the rest of the day. Workers in dept A help the injured worker – Lost production the rest of the day in dept B, and lost production in dept A.
Hour 6: The work orders for dept D are incorrect. It takes 20 minutes to get them corrected – Lost production in D.
Hour 7: A worker in dept C has a parent teacher conference and has to leave early. Lost producton in dept C for hours 7 & 8.
Hour 8: The workers in dept A decide to do a major clean up of the department spaces – subtract all production for hour 8 in dept A.
You can see that what management thought was a balanced process is anything but balanced in practice. There are just too many things going on to accurately predict the output of each department. This is why it is important to know which processes are constraining the output at any given time.
It is a very powerful concept to understand that the output of a bottleneck process determines the output of the entire plant or organization. Consider this simple financial example:
Let’s say that a small department of three people is the bottleneck of a factory that has 300 people and produces $100mil per year and could easily take another $10mil in orders if they had enough capacity. There really isn’t enough work to justify another full person in the bottleneck department because they are very highly skilled people and expensive. If another person was hired, the department would produce the additional 10% needed, but then all four people would be idle for 20% of their day. Many managers would balk at having four expensive people idle for this amount of time. But look at what it would mean to the plant. If the cost of the additional person was $100,000, this would be the investment needed to increase the plant output by $10mil. Let’s make the scenario a little more complicated and say that the value of what the three person department produces is less than what it costs the company in pay and expenses of the department. The value of the bottleneck should be viewed as the value produced by the whole organization when making decisions about that department. In this case the company produces approximately $500,000 per hour ($100mil/2,080hrs). This is the true value of the bottleneck’s production. When the expense of $100,000 over a full year to is compared to the additional production of $500,000 per hour for the plant, the decision to hire gets a little easier. Of course, another department may emerge as the next bottleneck before the entire $10mil in output is realized, but the fact remains that the additional expense will be small in comparison to the benefit.
Although it would be too time consuming to get into how to find a bottleneck in this post, particularly because there is almost always very different opinions in any team on where to find the bottleneck. The reasons provided are usually entirely anecdotal. This is a good place to start however. Try to get the team to agree on a few candidates for the bottleneck and then begin to collect data to find out which one is the biggest constraint according to the data collected.
Once you have identified what is believed to be the real bottleneck, what things can you do to make sure the output of that process is maximized. Here are some suggestions:
1. Add people
2. Move people from non bottleneck areas
3. Add Equipment
4. Reduce set up time
5. Farm out work to other departments
6. Farm out work to vendors
7. Train people to be more efficient
8. Move in more talented people
9. Speed up equipment
10. Move equipment from other areas to the bottleneck
11. Stage parts
12. Ensure quality
13. Redesign product
14. Improve process
15. Provide better supervision
When else should bottlenecks be considered?
1. Buying equipment – Don’t buy equipment for any other area if the bottleneck area needs anything.
2. Sick days – If someone calls in sick who works in the bottleneck area, take action immediately to backfill.
3. Starting & stopping times – Make sure that work is being performed every minute available, either through extra supervision, or other methods.
4. People coming in late or leaving early – make sure this doesn’t happen unless planned ahead with backup personnel to keep the work moving.
5. Maintenance – Make sure the machinery in the bottleneck is the most reliable.
6. Inventory (safety stock) – extra components may be wasteful if they are everywhere, but it will pay off in the bottleneck area.
7. Work order accuracy – Extra time checking for 100% accuracy is cost effective in the bottleneck area, but probably not factory wide.
8. Scheduling priority – Make sure anything needed by the bottleneck process has top priority if it will keep the area from losing time.
9. Better supervision – The bottleneck area should have the best available supervision.
10. Engineering Changes – Any changes from engineering to make the process more efficient in the bottleneck area should be implemented immediately, even if it is only an incremental improvement and large improvements could be implemented in other areas.
No comments:
Post a Comment