White Paper · Spring 2026

JOpt - The optimization engine that respects every constraint.

JOpt.TourOptimizer is a production-grade backend engine for route optimization, workforce scheduling, and resource planning — designed for the complex, conflicting constraints of real operations.

Explore Documentation → Talk to Us

10 min

To First Optimization

80+

Tunable Properties

REST

Any Language via OpenAPI

Docker

One-Command Deployment

The Problem

Shortest path is only the beginning.

Real dispatch demands balancing dozens of competing constraints — time windows, driver skills, load capacities, zone rules, SLAs, overtime, and CO₂ targets — simultaneously. Simple routing tools cannot handle this. Manual dispatching does not scale.

12+

Constraint types in a typical scenario

∞

Route permutations at 50+ stops

100+

Working Java examples included

Hard constraint violations — ever

"JOpt.TourOptimizer is not a single algorithm product. It is a feature-complete optimization platform shaped by customer scenarios across logistics, field service, manufacturing, and dispatch operations."

— Dr. Jens Richter, DNA Evolutions GmbH

What is JOpt.TourOptimizer

A backend optimization engine.
Not a frontend. Not a SaaS dashboard.

JOpt.TourOptimizer is the computational core that sits inside your product — integrable as a native Java library or as a containerized REST microservice accessible from any programming language.

How it works

You send nodes (tasks/stops), resources (vehicles/technicians), constraints, and objectives. The engine returns optimized, feasible schedules — respecting every rule you define.

Internally, JOpt uses a multi-phase pipeline: construction algorithms (savings heuristics, clustering, or quantum computing via D-Wave), simulated annealing, and a genetic algorithm with both random and intelligent operators.

The OpenAPI contract is derived directly from the internal snapshot model — generated clients are always aligned with the engine, and optimization states are portable and reproducible.

Two integration paths

Path	Technology	Best for
Java SDK	Maven / JAR	In-process, maximum control
Docker + REST	Spring WebFlux, OpenAPI	Any language, microservices

REST clients provided for:

Supporting products: RouterPlanner & GeoCoder

JOpt.TourOptimizer is the core engine. It is complemented by JOpt.RouterPlanner (realistic travel times via OSRM/Valhalla) and JOpt.GeoCoder (Pelias-based geocoding) — available as SaaS components to feed the optimizer with production-quality data. Contact us for access.

Key Differentiators

What makes JOpt.TourOptimizer different.

Every feature was developed in close collaboration with customers running production systems. These are solutions to real dispatch problems — not academic exercises.

🔒

Hard Constraints by Architecture

SLAs are guaranteed structurally — not approximated with penalties. Pillar nodes, relations, and zone constraints cut the solution space. Every solution is valid.

📦

First-Class Pickup & Delivery

Not an add-on. PND is a full feasibility system tracking goods in vehicles at all times — with timed loads, flexible quantities, manufacturing planning, and audit-grade reports.

🎯

AutoFilter — Scale Without Sacrifice

Structurally removes violation-prone nodes from the problem instance. Not high penalties — actual removal. Accelerates convergence in large, constraint-heavy scenarios.

🧩

Open Assessor — Your Rules, First-Class

Inject custom costs, violations, and domain logic at node or route level — without forking the solver. Your business rules become part of the engine.

🔄

Warm Start & Plan Continuity

Inject existing plans, re-optimize incrementally, insert late jobs, add or remove resources. Your world changes — JOpt adapts without starting from scratch.

⚛️

Quantum Computing Plugin

D-Wave quantum annealing for TSP construction. One of the first routing engines with production-ready quantum integration. Future-proof your optimization stack.

Capabilities

The full feature landscape.

14 capability categories — from time modeling to skill matching to CO₂ optimization — each with deep configuration and production-tested behavior.

Time & Schedule Modeling

Multi-day planning with overnight stay policies
Positive & negative FlexTime (flexible shift starts)
First/last node constraints for depot logic
Joint visit durations for co-located tasks
Wait-on-early-arrival behavior (global & per-node)
Resource visit duration efficiency factors
Event nodes for non-geographical tasks

Constraints & Hard Rules

Pillar/Captured nodes — guaranteed SLA matching
Relations — same route, same visitor, time-window coupling
Zone codes & zone crossing penalization
Bridge/tunnel crossing with asymmetric costs
Magneto conditions (soft node attraction/repulsion)
Open/closed routes, return-to-start, alternate destinations
Capacity plausibility checks on startup

Skills & Resource Matching

Mandatory, preferred, banned, unpreferred resources
Expertise levels with min/max thresholds
Three skill cost models (exact match, prefer high, no penalty)
BitSet-accelerated matching for large skill matrices
Time-dependent / expiring qualifications per shift
Resource constraint alias IDs across systems
Resource connection & visit duration efficiency

Performance & Optimization Control

Performance Mode — up to 50% faster optimization
Clustering construction for large instances
AutoFilter with selective violation categories
Algorithm selection pipeline (construction + heuristic phases)
Custom cost convergence criteria
CO₂ emission optimization as first-class objective
80+ tunable properties (weights, margins, thresholds)

Pickup & Delivery (PND)

Full goods-tracking feasibility system
SimpleLoad, TimedLoad, FlexLoad, UnloadAllLoad
Manufacturing planning with on-demand production
Capacity factors for heterogeneous cargo
Optional dump/reload/production infrastructure nodes
Fuzzy exchanges (partial fulfillment)
Audit-grade ResourceDepot & NodeDepot reports

Continuity & Observability

Save/load snapshots (JSON / JSON.BZ2)
Custom warm-start solutions from JSON
Fire-and-forget mode with MongoDB persistence
Real-time progress streaming via reactive events
Comparison tool for before/after analysis
KML export for GIS visualization
Deterministic exception handling for production services

View Complete Feature List →

Quick Start

First optimization in 10 minutes.

JOpt provides Docker-based sandbox environments with a browser IDE. No local Java, Maven, or toolchain required — just Docker and a browser.

1. Start a sandbox

docker run -it -d \
  --name jopt-examples \
  -p 127.0.0.1:8042:8080 \
  dnaevolutions/jopt_example_server:latest

2. Open your browser

Navigate to http://localhost:8042 — log in with password jopt. A full browser IDE opens with all Java examples ready to run and modify.

3. Run your first example

Open FirstOptimizationExample.java, click Run, and watch nodes get assigned to optimized routes. Modify constraints and properties — re-run to see the effect.

Available sandbox environments

Sandbox	Image	Port
Java SDK	`jopt_example_server`	8042
Java REST	`jopt_rest_example_server`	8043
Python REST	`jopt_py_example_server`	8033
C# / .NET	`jopt_net_example_server`	8023

Free evaluation license included

Optimize up to 20 elements without any license. Extended free license (7 resources + 30 nodes) available with sign-up.

Sandbox Quickstart Guide → First Optimization Tutorial →

Enterprise Deployment

Your infrastructure. Your rules.

JOpt.TourOptimizer is a stateless Spring Boot application — designed for horizontal scaling and external orchestration. DNA Evolutions provides prebuilt, versioned container images. You handle the infrastructure.

Deploy anywhere

docker run -d --rm \
  --name jopt-touroptimizer \
  -p 8081:8081 \
  -e SPRING_PROFILES_ACTIVE=cors \
  dnaevolutions/jopt_touroptimizer:latest

Swagger UI at localhost:8081/swagger-ui/index.html

🐧

Linux

🪟

Windows

🍎

macOS

☸️

Kubernetes

Terraform reference architectures available for AWS, Azure, GCP, and on-premise. No sticky sessions, no shared state, no custom build tooling required.

System Architecture Guide →

Industries

Proven across sectors.

JOpt's feature breadth adapts to your domain — not the other way around.

🚛

Logistics & Distribution

Fleet routing, PND, multi-depot, capacity management, last-mile

🔧

Field Service

Technician dispatch, skill matching, SLA guarantees, multi-day

🏭

Manufacturing

Production-integrated PND, supply chain, on-demand planning

🏥

Healthcare & Services

Home care, sample collection, appointment routing, strict SLAs

Path to Production

From evaluation to deployment.

Designed for progressive adoption — start small, validate your modeling, then scale.

Evaluate

Run a sandbox. Model 5 nodes and 1 resource. Understand nodes, time windows, connections, and results in 10 minutes.

Model

Add real constraints — skills, PND, zones, relations. Validate feasibility with small instances before scaling.

Integrate

Connect via REST or Java SDK. Use snapshot IO for persistence. Plug in your distance matrix.

Scale

Deploy on Kubernetes or Terraform. Horizontal scaling, fire-and-forget for long-running jobs.