QuSolve Solver to optimize production processes

Main article: Business Process Management

2024: OptJet Presentation

The Russian developer "Quantum Systems" (QuSolve) presents a completely import-independent OptJet solver designed to solve complex industrial problems. OptJet is a domestic alternative to foreign solver companies CPLEX and Gurobi that have left the Russian market. QuSolve announced this on October 22, 2024.

Most companies are faced with the need to solve complex problems of mathematical optimization in order to reduce costs or increase profitability in managing complex production, logistics and other business processes. For example, to create an optimal monthly production plan for a workshop with several alterations and a set of parallel units, a mathematical model with millions of variables may be required. At the same time, the term for finding a solution should be no more than a few hours - otherwise, by the time the result is received, the original data will have time to change so much that it will be useless.

Specialized software - the so-called "solvers" or "solvers" - can cope with tasks of this scale in conditions of such severe restrictions. Long-term leaders of this market were IBM (CPLEX solver) and Gurobi (solver under the same brand), however, in recent years, Chinese COPT and OptVerse solvers have appeared (the latter developed by Huawei), and recently a Russian solution has become available - OptJet.

Solver OptJet is developed based on the experience of solving real industry problems and taking into account the needs of Russian customers, has a separate module for creating and debugging an optimization model in the form of a Python library with a convenient interface, which uses its own special sub-language. Solver can solve tasks such as planning and graphics, including for industries that combine continuous and discrete units, scheduling and routes, dynamic pricing, placing objects in space, developing network topology, mixing material and energy balances, and many others.

It is worth noting that the successful implementation of solver at a large enterprise is not always possible in the out-of-the-box installation format ON - in situations where an improvement in the optimized indicator by only 1% creates an effect of hundreds of millions, rubles a deep business analysis that precedes the translation of the problem into the language of mathematics has an important role. The specialists of the QuSolve team, when implementing their solution, are ready to deeply immerse themselves in the business and technological processes of the customer in order to ensure seamless integration, adoption and, as a result, real economic effect.

Using OptJet solver, for example, we managed to develop an end-to-end solution for integrated production planning and logistics, adapted to the needs of various enterprises. In other words, application software can be developed on the basis of the product core to solve specialized industry problems. Conducting a business analysis, we pay special attention to complex cases, for example, resource conflicts or exceptions to the rules that are not reflected in the documentation, "said Dmitry Vasilkov, founder and CEO of QuSolve.

However, customers with experience in translating business problems into logical and mathematical form can effectively use the OptJet solver themselves. Solver has a rich set of settings, which allows you to repeatedly speed up the search for a solution using knowledge of the internal structure of the problem. To familiarize yourself with the solver, a demo version is available, limited to 100,000 variables. OptJet Solver supports operating systems Windows 7 and above, macOS 12 + based, with Intel Linux glibc version 2.27 and above, with Python version 3.9-3.11 installed. The features of the Russian solution include always available technical support.

Another important category of potential solver users are companies that develop their own software systems, such as ERP systems, designed to automate and integrate basic business processes, such as finance, logistics and human resources management, CAD (computer-aided design systems), which are used in the development of engineering projects, MES (production process management systems), which provide monitoring and optimization of production processes, and BI (business analytics systems). In most such systems, the optimization functionality, the core of which is the solver, plays an important role. Until recently, foreign solvers were embedded in such systems, and now it is possible to seamlessly import them.