TECHNICAL FIELD REPORT · 3D NESTING OPTIMIZATION
From geometry to production:
methods, testing and findings.
Nesting Solver combines multiple placement strategies on real production parts. The aim is to preserve quantities and manufacturing constraints while achieving lower build heights, better volume utilization and more balanced layer density.
Up to 30% improvement in build height. This figure describes the height improvement reported in the team’s current field evaluations.
01 / DATA AND SCOPE
Real parts, different production scenarios.
Development uses part geometries and order quantities from the production needs of different companies. Repeated orders involving the same parts in different quantities are considered alongside mixed plans containing many part types.
Aerospace and defense, automotive, technical maintenance and printing-center scenarios provide different conditions for geometry complexity, orientation requirements and part distributions. Customer identities and design files are not published in this report.
02 / ALGORITHMIC APPROACH
Multiple strategies, shared manufacturing constraints.
The search is not limited to a single placement order. Different part sequences, allowed orientations, candidate positions and restart strategies are compared. Longer searches evaluate thousands of candidate placements to explore better results.
Geometry-based candidate placement
Parts are evaluated through voxel-based volume representations. Candidates are filtered against build-space limits, part clearance and orientation constraints.
Multiple starts and local improvement
Different ordering and rotation preferences create alternative starting points. Local rearrangements reduce dependence on a single initial solution.
Chain prevention
Parts can become interlocked through looped geometries without direct contact. Chain-prevention checks evaluate relationships that may prevent parts from being separated after production and aim to reject risky candidates.
Fixed-height layer balancing
Part distribution is reassessed without increasing the selected height. The aim is to reduce density peaks and differences between layers. This separate step preserves the height search’s original resume checkpoint.
03 / TEST PROTOCOL
Keep comparison conditions consistent.
Placement comparisons retain the same parts, quantities, build area and spacing. This separates layout improvement from producing fewer parts or reducing required gaps.
- Input preparationDefine STL models, quantities, machine dimensions and orientation constraints.
- Reference layoutRecord the starting arrangement and its build height.
- Alternative searchesExplore different strategies and starts; retain discovered improvements.
- Geometric checksEvaluate collisions, clearance, boundaries and enabled chain-prevention constraints.
- Result analysisMeasure height improvement and review layer balancing separately.
04 / EVALUATION METRICS
Assess placement quality as well as height.
| Metric | What is evaluated | Objective |
|---|---|---|
| Build height | Highest extent of the placed parts | Lower production height |
| Volume utilization | How the build space is used | Reduce unused space |
| Geometric feasibility | Collisions, gaps and machine limits | Preserve defined constraints |
| Separability | Potentially interlocking part relationships | Prevent chains |
| Layer density | Peaks, distribution and layer-to-layer variation | More balanced distribution |
| Search progress | Valid improvements found over time | Comparable recorded results |
Layer-balancing analysis can consider the density profile, peak values, standard deviation and variance together. The aim is to smooth avoidable peaks as well as improve overall distribution.
05 / FIELD FINDINGS
Up to 30% improvement in build height.
The improvement reported in current field evaluations is up to 30% in build height relative to a reference layout. Geometry diversity, quantity distribution and allowed orientations affect which strategy performs best.
Lower placement creates an opportunity to plan fewer production layers. Chain prevention and layer balancing complement height search by addressing geometric and layer-distribution quality.
Source: the team’s development with real production geometries and its updated field observations. The 30% figure concerns build height, not a claimed cost or time percentage.