Project – Indoor Route Planning

Indoor Route Optimization & Milk-Run System for Manufacturing Facility

Context

A large manufacturing facility relied on manual planning for internal material delivery routes, leading to inefficiencies in tugger movement and inconsistent milk-run execution.

Problem

Routing inside the facility was constrained by:

  • Physical obstacles across the plant floor
  • One-way and two-way aisle restrictions
  • Complex pickup and drop-off relationships
  • Equipment limitations (tugger length, load pairing)

Existing planning methods could not account for all constraints simultaneously, resulting in suboptimal routes and wasted movement.

Approach

Developed a grid-based routing system that translated physical plant constraints into a computational model.

Key steps:

  • Created a spatial grid overlay within AutoCAD aligned to the facility layout
  • defined route constraints:
    • Obstacles (0)
    • Directional paths (north, south, east, west)
    • Two-way flows (NS / EW)
  • Built an engine using A* and Dijkstra algorithms to compute shortest paths between all points under real constraints
  • Generated a full distance matrix reflecting true navigable paths (not Euclidean shortcuts)
  • Mapped all line-side and warehouse locations to grid coordinates

Optimization layer:

  • Designed an user friendly route builder using the distance matrix
  • Applied linear optimization model to evaluate route combinations and minimize total travel distance
  • Incorporated operational constraints:
    • Tugger length limits
    • Mandatory item pairing rules
    • Pickup and drop-off sequencing
    • Mandatory pick up and drop off side (Right Hand, Left Hand)

Result

  • Established a constraint-aware routing system aligned with real plant conditions
  • Reduced unnecessary travel distance and routing inefficiencies
  • Increase route utilization capacity to near 94%
  • Achieved rapid adoption by designing for operator usability and iterating closely with floor teams

Tools & Methods

AutoCAD, Excel, Python (A*, Dijkstra), optimization modeling (Linear Optimization)

Notes

Details generalized due to confidentiality.

Contact Us