“We just need to run, and run without leakers” quipped one of the maintenance managers at Turkey Hill Dairy in the fall of 2010. Improving quality and increasing throughput are two things that are important in the production of dairy products for this Pennsylvania-based company.

Realizing there was an issue with leaky milk and juice containers on several of their lines, Turkey Hill contacted Balluff for help. One line was analyzed and the primary problem was identified – caps not being seated correctly at the capping machine. By either under- or over-tightening due to variations in the bottle material, some of the bottles became susceptible to leakage at any point in the warehousing or delivery stages.

The Challenge: Identifying leaky bottles

The “leakers” had a series of compounding effects. First, the milk or juice had to be cleaned up at the point where the leak took place. If the leak occurred in the dairy, the cost was minimal. However, if the leak happened at the end customer, the customer could back-charge Turkey Hill for the cleanup cost and return the product.

There were also “leakers” discovered in transit. Not only did this mean that the skid would be refused by the customer, but it would most certainly lead to a truck having to be cleaned. In the worst case, a truck could be taken out of service for up to 4 hours and entire skids of product would be thrown out.

Loss estimates as high as $5,000 for a single worst-case event and up to $30,000 per month were recorded by the management team at Turkey Hill. In addition, the re-work involved on the full and partial skids of the returned and contaminated bottles was as many as 4,000 bottles per day.

The Solution: Implementing vision inspection

After some careful thought and testing, the decision was made to add an inspection station near the end of the line with a Balluff Sharpshooter vision sensor. The goal was to improve quality by removing “leakers” before hitting the palletizer with a better system of product checking than having human operators inspect the bottles as they went by.

The Results:

Availability (50 min per hour – 1200 min per day ideal)
1148 min per day average before changes
1142 min per day average after changes
Uptime before = 95.7%
Uptime after = 95.2%

Throughput (156,000 bottles per day ideal)
140,712 bottles per day average before changes
144,453 bottles per day average after changes
Throughput before = 90.2%
Throughput after = 92.6%

Quality (100% ideal)
Good vs. ideal before = unknown (estimate was 99.6%)
Good vs. ideal after = 99.5%

Total OEE before = .957 x .902 x .996 = .857 = 86.0%
Total OEE after = .952 x .926 x .995 = .877 = 87.7%

NOTE: The line is designed to run 130 half gallon drink bottles per minute, up to 3 shifts per day.

In the end, the availability of the line was reduced by 0.5% due to line stoppages to remove the leakers, but total OEE was increased by 1.7%. How? By catching more of the potential leakers before they left the dairy, the production time increased 2.4% when allocation was no longer needed to make up for losses or “re-work”.

As a side benefit, Turkey Hill was able to reduce waste and avoid penalties with their customers. The potential in over $300,000 in annual savings easily paid for the installation of the new reject station within the first two weeks. Also, the labor previously allocated to cleanup and re-trucking was converted to production and process improvement activities elsewhere in the dairy.

Making the most of your production line

The simple addition of an inspection station and detection of potentially “bad” products is a great example of how this customer reaped the benefits of a huge cost savings. In the future, Turkey Hill hopes to follow the same kind of methodology to bring improvement to other lines in their dairy. The Engineering Manager at Turkey Hill explained, “The increase in production by almost 2% helps us sustain our profitability.”