Roadmap

Last updated: 2026-02-23

This roadmap outlines major initiatives we are actively pursuing to push operational optimization beyond today’s standard tooling.
It is a forward-looking document: items may evolve as we validate feasibility, customer value, and production readiness.

Interested in joining an early access program or co-designing a feature? Contact us

Guiding principles

• Real operations first: every roadmap item must solve a concrete operational pain point.
• Compatibility by design: new capabilities should integrate with existing JOpt SDK and REST workflows.
• Measurable outcomes: runtime, feasibility, acceptance, and KPI impact must be quantifiable.
• Enterprise readiness: security, reproducibility, and deployment options are considered from day one.

---

Quantum-assisted construction (D-Wave)

What: Use a D-Wave quantum computer for parts of construction / initial solution building.
Why: Construction quality strongly influences downstream improvement phases. Better starting solutions can reduce runtime and improve final KPI quality.

Planned product direction

• “Quantum-assisted construction” as an optional construction backend
• Feature flags / safe fallbacks to classical construction
• Benchmarking framework to quantify value per scenario type

Potential impact

• Faster time-to-first-feasible schedule in some classes of instances
• Improved solution quality for highly constrained inputs

Status: Prototype demonstrated

• QUBO formulations created
• D-Wave integration proven as a plugin for the TourOptimizer Spring REST application

---

IoT platform with optimizer functionality (OpenRemote-based)

What: Provide an IoT platform with built-in optimizer integration, based on OpenRemote.
Why: Many customers want optimization directly connected to live operational signals: telemetry, device states, logistics, assets, and real-time constraints.

Delivery model

• Self-hosted (customer-controlled tenant)
• Managed service (DNA-hosted SaaS)

Core capabilities (planned)

• device and asset modeling (vehicles, containers, machines, tools, sensors)
• event-driven optimization triggers (e.g., threshold exceeded, device alarm, schedule drift)
• real-time dashboards + audit trails for decisions
• integrations into existing stacks (REST, webhooks, connectors)

Expected impact

• “Closed loop” operations: sense → decide → optimize → execute
• Reduced manual coordination between IoT data and planning systems

Status: Prototype demonstrated

---

AI for planning input, optimization quality, and explainability (end-to-end AI layer)

Vision: Build an AI layer that makes optimization easier to configure, faster and higher quality, and easier to trust — while maintaining enterprise requirements like governance, reproducibility, and auditability.

This initiative combines three capabilities that belong together in real projects:

A) Natural-language input → structured planning intent

What: Allow users to express intent in natural language, for example:

• “Try to keep these customers together”
• “Prefer early visits for VIP accounts”
• “Avoid risky combinations”
• “Explain why this route looks like that”

Planned capabilities

• intent-to-constraint suggestions (hard vs soft)
• interactive refinement (“Did you mean a hard rule or a preference?”)
• validated output mapped to known JOpt constructs

Expected impact

• faster configuration, fewer misunderstandings
• smoother onboarding and better usability for planners

B) Learning-enhanced optimization (pattern detection + reinforcement learning)

What: Use AI to improve optimization outcomes by detecting patterns and learning which moves/settings work best for specific scenario classes.

Planned product direction

• reinforcement learning and learning-guided move selection
• adaptive heuristics per scenario type (learn what matters in your domain)
• recommendations for run settings (properties/tuning) based on instance features
• strict reproducibility options (controlled randomness, auditability)
• evaluation harness to measure KPI impact vs runtime

Expected impact

• shorter runs (time-to-good solution)
• higher quality and more stable results across days
• better default settings without manual tuning

C) Explainability and governance (AI + rule engine)

What: Help users and stakeholders understand results and changes by adding an explanation and governance layer.

Planned capabilities

• decision-maker summaries (“what improved / what changed / why”)
• anomaly detection (“why is this route unusually long?”)
• constraint-driver explanations (“what forced this assignment?”)
• scenario comparison and drift analysis
• policy checks and compliance reporting

Expected impact

• higher acceptance by planners and end users
• faster debugging and safer rollouts
• stronger audit trails for governance-heavy environments