Operations Research 2025
Abstract Submission

177. Inspector Scheduling for German Road Inspection

Invited abstract in session TA-1: Routing, transportation, and scheduling, stream Mobility, Transportation, and Traffic.

Thursday, 8:45-10:15
Room: Audimax

Authors (first author is the speaker)

1. Patricia Ebert
Zuse Institute Berlin
2. Thomas Schlechte
, LBW Optimization GmbH
3. Stephan Schwartz
Zuse Institute Berlin

Abstract

Every year, over 500,000 truck and bus inspections are carried out by road inspection services of the German Federal Logistics and Mobility Office (BALM). The controls are performed by teams of inspectors and can be mobile (traversing the road network) or stationary (e.g., at certain service areas). Furthermore, a number of control topics must be assigned to a duty in order to specify the focus of the control and to achieve the road inspection control targets of the BALM.
We present a model to solve the respective duty scheduling and crew rostering problem in order to obtain duty rosters that comply with numerous legal regulations while maximizing the 'control success' in order to achieve the control targets. Extending an existing model for the scheduling of toll enforcement inspectors, we formulate the Template Assignment Problem (TAP), where feasible combinations of control topics (called templates) are assigned to the duties from the basic model. The TAP is modeled as a mixed-integer linear program on a directed hypergraph. The hyperarc approach allows us to easily model interdependencies between single assignments. For instance, certain combinations of control topics are particularly desired or undesired to be processed in parallel within a single control.
Moreover, many constraints regarding inspector's qualifications and the allocation of control topics can be satisfied during the construction of duty templates and hyperarcs.
This allows to reduce the complexity of the final optimization problem.
Since our approach is used in production by BALM to create the inspectors' duty rosters, we prove the effectiveness of our model on a number of real-world instances.

Keywords

Status: accepted

Back to the list of papers