Aqarios’ platform Luna v1.0 marks a serious milestone in quantum optimization. This launch considerably improves usability, efficiency, and real-world applicability by introducing FlexQAOA, a hybrid quantum algorithm designed particularly to deal with industrial constraints straight inside quantum circuits.

Luna allows professionals throughout logistics, power, manufacturing and different industries to simply mannequin, clear up, and interpret complicated optimization issues, bringing quantum computing into sensible use immediately.

On June 4, 2025, we launched Luna v1.0, our most important improve to our quantum optimization platform but. Quantum computing guarantees highly effective optimization capabilities, however the complexity of actual world issues, particularly dealing with downside constraints, nonetheless poses a serious problem on this subject. Luna v1.0 tackles these challenges head-on.

What makes Luna v1.0 particular? It supplies an intuitive person expertise, deeper integration of quantum and hybrid algorithms, and native assist for FlexQAOA, our new constraint-native quantum optimization algorithm.

Luna is designed for professionals who face complicated issues in decision-making. It does not require prior quantum computing experience, simply real-world optimization challenges that demand sturdy, constraint-respecting options.

Quantum optimization meets constraints: Why we constructed FlexQAOA

Quantum computing guarantees important breakthroughs in optimization, however when real-world constraints are launched, many quantum strategies falter. Industrial issues not often exist with out constraints: budgets, timeframes, capacities. Conventional quantum algorithms can’t deal with these constraints natively, requiring difficult downside reformulations by way of penalty phrases.

We noticed a chance: Why not create a quantum algorithm constructed particularly round constraints from the beginning?

That is FlexQAOA.

Constraints are in all places

Constraints are basic to optimization issues and outline the constraints of which options are legitimate within the first place. Contemplate the next situations:

• Power grids: Balancing energy era inside strict regulatory and capability limits.
• Manufacturing: Effectively assigning duties to machines whereas respecting capacities and deadlines.
• Logistics: Optimizing automobile routes with exact load limits and supply home windows.

Conventional quantum strategies battle right here. The reformulations to work with constraints waste sources exploring infeasible options, complicate the optimization course of by having to stability the unique goal with extra penalties, and require extra slack variables, growing the variety of required qubits.

We aimed for one thing easier, clearer, and extra environment friendly.

How FlexQAOA solves constraints straight

FlexQAOA straight encodes constraints into quantum circuits, eliminating difficult penalty buildings. We launched two ideas:

• XY-Mixers: Deal with choices requiring unique (one-hot) picks by constraining the quantum state to possible options.
• Indicator Features: Handle inequality constraints (like budgets or capacities) by making use of focused part shifts, effectively encoding constraint satisfaction.

These improvements enable FlexQAOA to ship clear, sensible options sooner.

FlexQAOA in observe: Benchmarking outcomes

We benchmarked FlexQAOA utilizing the multi-dimensional knapsack downside, a widely known, complicated optimization problem involving a number of constraints. Even at a couple of QAOA layers, FlexQAOA matches or surpasses baseline quantum strategies. With extra QAOA iterations, it persistently delivers superior outcomes, clearly outperforming standard penalty-based algorithms.

FlexQAOA achieves a likelihood of sampling high-quality options of greater than 90% at simply 10 QAOA layers for the investigated cases, with out requiring slack variables or tuning penalty weights.

Its constraint-aware structure enhances answer high quality and improves time-to-solution, because of a dramatically diminished search house, making FlexQAOA a powerful candidate for fixing industrial-scale issues as quantum {hardware} continues to evolve. The diminished search house not solely will increase efficiency but additionally allows the simulation of downside sizes which might be inaccessible to traditional strategies.

Detailed outcomes can be found in our current paper posted to the arXiv preprint server.

Actual-world impression: FlexQAOA in power optimization with E.ON

One of many first real-world functions of FlexQAOA was developed in collaboration with E.ON Digital Know-how, the place we addressed a key problem in the way forward for power: optimizing electrical energy demand from versatile home equipment in prosumer households.

The aim was to coordinate good units like EV chargers and warmth pumps in a means that minimizes electrical energy prices whereas making higher use of domestically generated renewable power—all with out violating grid constraints.

Utilizing FlexQAOA, we efficiently encoded the issue’s complicated construction straight right into a quantum circuit, enabling constraint-aware optimization that respects real-world feasibility. The outcomes present clear potential for bettering flexibility and effectivity in power methods.

You may learn the total case examine right here.

Luna v1.0: Quantum optimization for everybody

With Luna, customers can mannequin, benchmark and clear up optimization issues intuitively utilizing Python, whereas Luna supplies hardware-agnostic entry to varied algorithms and {hardware} backends to select from. It combines proprietary algorithms with automated pipelines, making the method of problem-solving extra intuitive and simpler than ever earlier than. Already in lively use throughout logistics, manufacturing, and power methods, Luna proves that quantum optimization is now inside attain for anybody seeking to get began.

Our roadmap consists of extending FlexQAOA for broader constraint varieties, enhancing efficiency on quantum {hardware}, and increasing hybrid optimization workflows. We imagine that is only the start of the transformative potential of quantum optimization.

This story is a part of Science X Dialog, the place researchers can report findings from their printed analysis articles. Go to this web page for details about Science X Dialog and tips on how to take part.

Aqarios is a Munich-based deep-tech startup, based in 2021 as a spin-off from LMU Munich and its “Quantum Purposes and Analysis Laboratory” (QAR-Lab). We specialise in advancing quantum computing for industrial use by reworking cutting-edge analysis into sensible functions. We concentrate on quantum algorithms and quantum-enhanced machine studying, making them accessible and usable throughout real-world situations.

With a powerful basis in sectors similar to aerospace, automotive, finance, power, logistics, and manufacturing, Aqarios leverages over a decade of quantum utility analysis to ship progressive, user-centric software program options. Our intuitive instruments present streamlined entry to quantum functions, algorithms, and {hardware}—empowering everybody from novice customers to seasoned consultants to unravel complicated issues extra effectively and powerfully.