Our objective was to evaluate the throughput and productivity performance of the U.S. Postal Service’s 33 deployed Small Package Sorting System (SPSS) machines.
Since 2015, the Postal Service has deployed 33 SPSS machines costing over $141 million. The SPSS machine was designed to provide automated package sorting capability, alleviate existing processing capacity shortfalls, and reduce manual sorting to support package delivery. Currently, the Postal Service is investing an additional $23 million to purchase seven more SPSS machines, scheduled to be operational in November 2017.
What the OIG Found
We found that on average nationally, the SPSS machine throughput performance goal was exceeded by about five percent from January 1, 2016, through July 31, 2017. The throughput was based on the average number of packages sorted by the SPSS in an hour. The throughput goal was 4,500 packages per hour, and the achieved throughput was 4,737 packages per hour. However, only 23 of the 33 SPSS machines, about 70 percent, exceeded the goal, and the other 10 were below the goal, from about 132 to 878 packages per hour.
We also found that on average nationally, the Postal Service was not meeting its SPSS productivity goal by about 17 percent from January 1, 2016, through July 31, 2017. The productivity was based on the average number of packages sorted by SPSS compared to employee workhours used to staff the SPSS machines. The SPSS productivity goal was 385 packages per hour and the achieved productivity was 319 packages per hour. Twenty-nine of the 33 SPSS machines, or about 88 percent, failed to meet the goal, and the other four were above the goal, from about 14 to 307 packages per workhour.
We conducted SPSS site observations from May to August 2017 at one high-performing site — the Atlanta, GA, Processing & Distribution Center (P&DC) — and four low-performing sites — Merrifield, VA, Richmond, VA, Mid-Carolinas, NC, and the Rochester, NY, P&DCs. We also reviewed and evaluated the Postal Service’s April 2017 Lean Six Sigma (LSS) SPSS project documents at the Columbus, OH, P&DC.
During our site visit to the Atlanta P&DC, we observed SPSS operations and compared them to the best practices identified in the Columbus LSS project. We observed best practices related to supervision and planning that included:
- Monitoring and correcting staff labor code selections for reporting workhours;
- Using standard work instructions for machine set-up and restart; and labor code selection; and
- Matching SPSS staffing to package volume.
As a result, SPSS productivity at the Atlanta P&DC averaged 486 packages per workhour during the period of review, exceeding the productivity goal by101 packages per workhour, or 26 percent.
At the four low-performing sites we observed lack of supervisory presence and planning that resulted in:
- Incorrect labor code usage for reporting workhours;
- Lack of standard work instructions for machine set-up and restarts and labor code selection; and
- Insufficient package volume to support SPSS staffing.
As a result, SPSS productivity was misstated and throughput did not meet the goals. Better supervision and planning will improve SPSS productivity and throughput.
We calculated that the Postal Service would save about $24.8 million in labor costs annually by correcting the causes of low productivity nationally. This will reduce costs, increase operational savings, and support the Postal Service’s package platform strategy.
What the OIG Recommended
We recommended management:
- Ensure adequate supervisor and staff SPSS training that includes standard SPSS machine set-up and restart instructions;
- Ensure staffing to package volume management;
- Ensure monitoring and correct staff labor code selection for reporting workhours; and
- Determine the nationwide applicability of the Columbus, OH, LSS project.