Invented by Olbrot; Lukasz, Gomolka; Zbigniew, Twarog; Boguslaw, Mysliwiec; Kamil, Golec; Karol, FIBRAIN spolka z ograniczona odpowiedzialnoscia

In this blog, we’ll explore a new patent on a method for optimizing process tasks, especially for choosing raw materials and semi-finished products in factories. The patent uses something called ALMM-Optim, which stands for Algebraic-Logical Meta-Model Optimization. You’ll learn why it matters, what came before, and how this invention works in a very simple and engaging way.

Background and Market Context

Factories and industries today are always looking for ways to use resources better, make less waste, and save money. One of the hardest things is deciding which raw materials or parts to use to make products, especially when there are lots of options and rules. For example, a company making fiber optic cables has to pick and use different spools of fiber, making sure they waste as little as possible and still fill every order.

Making these choices is not easy. There can be hundreds or thousands of ways to pick and use materials, but only some ways are allowed or work best. Every wrong choice can mean wasted materials, extra costs, or unhappy customers. So, companies want smart ways to help them decide what to do at every step.

Before, people often used simple rules, or basic computer programs, to help with these choices. But as factories grew and orders got more complex, these old ways could not keep up. They were too slow, too simple, or could not handle all the rules and changes that happen in real life.

This is where the new patent comes in. ALMM-Optim is a new system that helps factories make these choices in a much smarter way. It uses a special way to describe problems and solutions, making it easy to update, adapt, and get better over time. Unlike old systems, it can keep learning and store new ways to solve problems, which helps it handle even really hard or new tasks.

In the market, this means companies using ALMM-Optim can be more flexible. They can handle sudden changes, new orders, or new rules without starting over. They can also use less material, waste less, and save money. This is very important in industries like cable making, car manufacturing, and many others where choosing and using materials is a big part of the job.

Scientific Rationale and Prior Art

To understand why this invention matters, it helps to know what came before. Factories have long used math and logic to help with choices. The main ways have been mathematical programming (like linear programming) or logic-based methods (like constraint logic programming). These methods try to find the best way to do things, given a lot of rules and choices. But as problems get more complex, these methods can become slow or too hard to update.

For example, let’s say you need to pick spools of fiber to make cables, and you want to waste as little as possible. You have rules about how much you can cut, what orders you need to fill, and how much each spool costs. Older methods might look at every possible way to cut and choose spools, but this can take too long if there are lots of options. Or, they might use simple shortcuts, but these don’t always give the best answer.

Some older systems tried to store common problems and solutions so they could reuse them. This helped, but it was hard to keep these systems up to date, or to let them learn new kinds of problems. Many times, if the problem changed even a little, the whole system had to be changed or rebuilt.

Other approaches used decision trees or flowcharts, but these were not flexible. They worked only for the problems they were built for, and could not adapt on the fly.

The new patent moves past these old ways. It uses something called an algebraic-logical meta-model (ALMM), which is a fancy way of saying it describes problems and solutions using math and logic in a flexible way. This means it can handle lots of different kinds of problems, can be updated easily, and can even “learn” by storing new ways to solve tasks as they come up.

What makes ALMM-Optim special is that it does not just solve a problem once. It keeps a library or “knowledge base” of models, examples, and rules that grow over time. If a new problem is a lot like an old one, it can use what it learned before. If it’s new, it can use pieces from old problems to help build a solution. This is a big step forward over prior art, where systems were stuck with what they knew when they were built.

Another key idea is that ALMM-Optim handles problems as a series of steps or decisions. Each step depends on the current state—what materials are left, what orders are left to fill, what has been done so far. At each step, the system checks what choices are possible, what rules apply, and picks the best next move. If something changes—like a new order comes in, or a machine breaks—the system can adjust its plan right away.

This science-based approach helps solve what are called “NP-Hard” problems. These are problems where there are so many ways to do things that it’s almost impossible to check them all, even with fast computers. ALMM-Optim uses smart ways to cut down the number of things it needs to check, so it can find good answers quickly, even for very hard problems.

Invention Description and Key Innovations

Now let’s look closely at what ALMM-Optim does and how it works. The system is built around two main parts, or modules.

The first module is like a smart library. It stores models of different problems, templates for new problems, and ways to bring in data. This module lets users describe their problems using simple math and logic terms, so the system knows exactly what the problem is, what rules apply, and what the goal is. For example, in a factory, you might describe how many spools you have, how long each is, what orders you need to fill, and how much waste you want to avoid.

The second module is the problem-solving brain. It has two parts: a problem library and a set of rules about what makes a good or bad solution. The problem library holds both algebraic-logical models and software models for each problem. It also keeps a set of building blocks that can be used to make new models, and a way to keep track of what properties or rules go with which problems.

When you use ALMM-Optim, here’s how it works:

First, you put in your data—what you have, what you need, and any limits or rules. The system sets up the problem using its library, or lets you build a new problem if it’s something new.

Second, the system figures out where to start (the initial state). It checks what choices are possible at this point, what the rules are, and what the “goal” looks like (for example, all orders filled, or waste minimized).

At each step, the system:

• Checks what choices are possible now
• Picks the best possible move
• Moves to the new state that results from that move
• Checks if the new state breaks any rules or hits the goal
• If not done, repeats the process

One big innovation is how the system keeps track of all the possible states and choices, without getting lost. It defines every state using vectors (just a list of things like how much you have left, what time it is, etc.). Each choice you make leads to a new state. The system only looks at choices that are allowed, and keeps track of which states are “inadmissible” (not allowed) or “goal” (done).

Another smart idea is that the system can work with both “finite” (ending) and “infinite” (never-ending) sequences of actions. For most factory problems, you want a finite sequence: you start with some materials and end when all orders are filled or materials run out. But in some cases, you might want a system that keeps working as new orders come in.

The ALMM-Optim also handles something called “trajectories.” This means it can look at a path of decisions from start to finish, and figure out which path gives the best result. If something changes along the way, it can change the path without starting over. This is important for real factories, where things change all the time.

The knowledge base is another key part. Every time the system solves a problem, it saves what it learned. Next time, it can use that knowledge to solve new problems faster. As more problems are solved, the system gets smarter, helping factories become more efficient over time.

The system is also built to work for many users at once (multi-user) and can run on a client-server setup. This means big factories with lots of users can all use the same system, share knowledge, and keep improving.

Let’s look at a simple example from the patent. Suppose you need to pick a set of fiber spools to make cables, trying to waste as little as possible. The system starts with all spools unused and no cable made. At each step, it picks the next best spool to use, checks if this fills an order or creates waste, and keeps going. If it finds a way that fills all orders with the least waste, it saves this as a “goal state.” If it hits a dead end (no more good choices), it marks this as “inadmissible” and tries a different path. Over time, it gets better at finding the best way to use spools for any set of orders.

The user interface is made simple. You enter your problem using the terms and templates the system knows. You can use models from the library or build your own. The system helps you pick the right solving method and shows you the result—either the best solution found, or a message if no solution is possible. You can also see details of how the solution was found.

Another new feature is how the system can use different “criterion properties” to pick solutions. For example, it can handle problems where you want to add up scores (additive), multiply them (multiplicative), or use other rules. This makes it flexible for many types of problems.

In summary, the key innovations of ALMM-Optim are:

• A flexible, growing library of models and rules
• Smart handling of steps, choices, and states
• Ability to learn from past problems and get smarter
• Easy to update and adapt for new or changing problems
• Support for many users and big factories

All these features help factories use their materials better, waste less, and save more money, while being able to handle new challenges as they come up.

Conclusion

The ALMM-Optim patent brings a new way for factories to solve tough problems about picking and using raw materials. By using a smart and flexible system that grows and learns, it lets companies handle complex choices with ease. This means less waste, more savings, and the ability to adapt to new problems fast.

If you run a factory or work in a place where choosing what to use and when is a big deal, ALMM-Optim can help you do it better. It’s easy to use, keeps getting smarter, and works for a wide range of problems. With this technology, you can make better decisions, keep up with changes, and stay ahead in a fast-moving world.

Click here https://ppubs.uspto.gov/pubwebapp/ and search 20250217432.