Case studies

Swiss Post

 

Postal organization streamlined its parcel sorting system, improving address-recognition rates and reducing number of errors – resulting in more efficient handling of parcel traffic and decreased costs    
Postal organization streamlined its parcel sorting system, improving address-recognition rates and reducing number of errors – resulting in more efficient handling of parcel traffic and decreased costs
   

Business impact

Significant improvement in parcel sorting system, reducing video coding load by more than 25 percent and manual coding by approximately 15 percent. Taking into account Swiss Post's volume of parcels, the improvement is equivalent to a yearly volume of around 75 standard Swiss Post railroad trains with 10 cars each, whose contents would have been manually coded in the past. In financial terms, the productivity gain returned the initial investment in just a few months.

Issue

Swiss Post needed to improve its automated parcel coding process and wanted to streamline traffic within the parcel sorting centers for more efficient, cost-effective mail handling with minimal errors.

Executive summary

The IBM Research team identified an opportunity to optimize the automated parcel coding system and approached the client through IBM Global Business Services. Swiss Post decided to engage IBM to develop system enhancements that would improve its address-recognition rates to increase the number of automatically coded parcels. The result was more efficient organization of parcel traffic within the sorting centers. With the enhancements, video and manual parcel coding was reduced, and Swiss Post realized a significant cost reduction.

What IBM did

The Image Processing Group from IBM’s Haifa Research Lab worked with the Postal Solutions Center in Boeblingen and Swiss Global Business Services consultants to develop a solution that enhanced address block location, character recognition and logical post-processing algorithms. Its aim was to increase the recognition rate while reducing the number of substitution errors. Automatic recognition rates rose by about 7.5 percent to about 75 percent, and the number of errors was reduced by approximately 23 percent.

Capabilities applied

IBM Research Services the partnership between IBM Research and IBM Global Business Services, helps clients transform to an on demand business model that responds to changing needs by analyzing their business challenges and developing innovative ways to address them. By bringing together a team of IBM Research experts and IBM Global Business Services consultants, IBM Research Services helped Swiss Post improve its parcel sorting system, which saved the company money and time. The Haifa Research Lab provided a unique perspective, which was gained from the work it did that resulted in 30 patents in the image processing area as well as a solution it developed for error-free data entry.

Related case studies

More on the issue

Improving operational resiliency and efficiency

With business pressure more intense than ever and emerging solutions making it practical to contemplate substantive change, business has reached a tipping point.