BE1030565B1 - Computer program and method for dynamic route optimization using multiple artificial intelligence - Google Patents

Abstract

The present invention relates to a computer program and a method for finding a route for a road user in a traffic system, in particular the fastest possible route with a short traversion time, in particular a short travel time, for the road user through the traffic system, from a starting point in the system to a destination point in the system with road users using big data, comprising calculating a current ideal speed as a speed recommendation for driving along the route, in particular based on a current and / or expected traffic light switching at an upcoming traffic junction, in particular an upcoming intersection, as well as a density , direction and speed, in particular directional density and speed distribution, of road users in the side streets of the route, in particular the side streets that have not yet been passed in the direction of travel to an upcoming traffic junction, in particular an upcoming intersection, based on the real-time data. Furthermore, it is planned to revise the speed recommendation and re-examine the route suggestion if a revision condition occurs.

Description

RE oge] SYAL 1 BE2022/5412

Computer program and method for dynamic route optimization

Use of multiple artificial intelligence - patent application -

Field of invention

The present invention relates to the technical field of route and

Path optimization, with numerous different areas of application, for example

Navigation systems for vehicles, autonomous driving, semi-autonomous driving,

Driver assistance systems, drone flights or delivery robots. The invention can be used in particular in heterogeneous logistics and transport systems. at

2 BE2022/5412

Technical background

Finding an optimal route is a very old and yet still very modern task

Problem. Since the 1930s, numerous scientists have been working on the so-called “traveling salesman problem”, whereby a route with n

Stopovers had to be optimized in relation to their route. The behavior of others

Road users were regularly neglected in such calculations and considerations. Traffic regulations such as maximum speeds were also negligible. The time factor already played a minor role compared to the distance factor.

Navigation systems are used in numerous vehicles today.

A common problem faced by contemporary road users is that they have to get from a starting point to a destination point as quickly as possible

transport system.

In these systems there are problems with multiple participants blocking each other. For example, traffic jams arise from motor vehicles.

Modem mapping and navigation systems, e.g. the company's proprietary “Maps”.

Google/Alphabet Inc. are trying to remedy the situation through various mechanisms.

For example, current traffic information is taken into account, such as a

Occurrence of accidents, construction sites or traffic jams. For example, if a traffic jam occurs, an alternative route to a possible avoidance of the obstacle can be suggested to a road user.

However, a common problem is that, for example, a certain or uncertain

A while later a traffic jam occurs on the diversion route because many vehicles are diverted

Ee and

3 BE2022/5412 and the overall problem of the transport system is not addressed dynamically enough.

The approach is also not holistic enough.

Above all, the mutual blocking of road users of the same and/or a different type is a central problem in contemporary traffic technology with many

participants.

The present invention, which is presented below, is intended to provide a remedy and to create a generally applicable, efficient system with which the most suitable

Routes for crossing the system between the start and destination points can be created dynamically and dynamic adjustments can be made.

The invention should be applicable ubiquitously and across fields, while also being easily scalable and easily accessible, including, for example, in the large number of existing ones

Vehicles so that everyone can benefit from them.

The disadvantages of the prior art are overcome by the method

Claim 1 for finding a route through the transport system for one

Road users in this same area, as well as the independent claims of the others

Claim categories.

The invention thus contributes significantly to technical progress in dynamic

Route optimization using artificial intelligence. a | Ee Ee Ee

4 BE2022/5412

Description of the invention

The present invention provides a computer program with the features of

Claim 1.

Further advantageous refinements are specified in the subclaims.

Accordingly, a method for finding a route for a road user is in one

Transport system, in particular the fastest possible route with a short one

Traversion time, in particular a short travel time, for the road user through the traffic system, from a starting point in the system to a destination point in the system with road users using big data, which is the following

Steps include: Providing real-time data relating to other road users who are in the traffic system, in particular position and

Speed data, in particular data from terrestrial, locally distributed sensors,

Providing historical data on road users in the transport system

Routes traveled in the past, including the success factors assigned to them, which allow a comparison of the actual routes traveled with a respective original one

Quantify prediction and using a second method of artificial

Intelligence were generated, providing several route suggestions as a preselection for a route between the starting and destination points for the road user and evaluating these route suggestions using a first artificial intelligence method based on the real-time data and the historical data including success factors, with a temporal

Development of road users along the route is at least partially simulated,

Selecting the best-rated route, calculating a current ideal speed as a speed recommendation for driving along the route, in particular based on a current and/or expected traffic light switching at an upcoming traffic light

traffic junction, especially an upcoming intersection, as well as a density,

Direction and speed, especially directional density and

Speed distribution, of road users in the side streets of the route, 5 in particular the side streets that have not yet been passed in the direction of travel to an upcoming traffic junction, in particular an upcoming intersection, based on the

Real-time data, revise the speed recommendation and recheck the

Route suggestion when a revision condition occurs.

In one step, real-time data relating to other road users who are in the traffic system, in particular position and speed data, is provided. For example, these are vehicle positions or

speeds. The real-time data can also, for example

Environmental factors include, for example, the presence and extent of a

Construction site. Weather data can be another example.

The data is recorded in particular by terrestrial, locally distributed sensors. This

The system can always provide the most current and precise data at all installation locations. However, other variants can also be considered as embodiments, for example flying sensors on flying objects, such as flying drones.

Historical data is also provided. The historical data can, for example, include real-time data from the past or data derived therefrom

Data. In particular, however, the historical data can contain data about already from

Routes traveled by road users in the past. Here you can

Success factors are included, which are assigned by an AI (the second AI). The second AI can be part of a control loop or self-regulating

6 BE2022/5412

Feedback loops show how will be described in more detail later.

Success factors quantify, for example, the extent to which a route prediction corresponds to the actual time course of a route traveled. A non-limiting one

An example of a success factor is a weight in the form of a floating point number between 0 and 1, where 1 quantifies exact agreement with the prediction. However, this model can be extended so that success factors greater than 1 are possible, for example if a route was actually completed even faster and more efficiently than predicted.

More complicated data structures can also be used as success factors.

First, several route suggestions are provided. This can be one

be pre-selected. For example, such suggestions are generated by well-known methods. External systems that are already available can also be used (for example, well-known Maps services, such as Google Maps).

navigation systems, or similar). A temporal development/prediction is carried out. Through the first AI, which also takes part of the...

Control loop/control loop, the route suggestions are evaluated in order to select a best suggestion (e.g. highest rating).

A speed recommendation is provided in the form of a current ideal speed. Such a speed recommendation can be handled in different ways: For example, in autonomous driving, an attempt is made to keep the vehicle as high as possible

based on this speed. With a navigation system, the speed recommendation would be displayed to a driver. It can also be, for example

Cruise control system (e.g. cruise control) can be controlled.

The following factors are particularly taken into account when calculating the instantaneous ideal speed: a Se

7 BE2022/5412 a) a current and/or expected traffic light switching on an upcoming

Traffic junction, especially an upcoming intersection. Traffic sign systems such as traffic lights are often used on a recurring basis, but are flexibly adapted to circumstances

Rhythms switched. Various techniques for consideration come into play

Mission. For example, the data according to which the traffic light is switched is tapped and used (abstract rule data). Alternatively, current data is used

Traffic light switching, for example provided by the traffic light switching (instantaneous data, for example red/yellow/green) and/or a (currently remaining) time is provided until the next switching process. Furthermore, the traffic lights can also be switched externally

Sensors are monitored. This makes the current traffic light switching accessible. However, the rules for switching traffic lights can also be derived from the switching behavior of the traffic lights over a certain period of time. For example, a (particularly third) AI is used for this purpose.

Combinations of these options are possible and increase data consistency.

A traffic light or traffic sign system or traffic control system is every dynamic one

Controlling a traffic flow at a traffic junction. A color representation, as we know it from road traffic, is just an example.

By taking this data into account when recommending speed,

Vehicle - to put it very loosely - is stopped using technical means to "drive with foresight" - in such a foresighted form that the most foresighted driver could never do himself due to a lack of information.

A positive effect on the overall traffic dynamics has already been proven in simulations when a proportion of approximately 8 - 10% of road users already use this aspect of the method according to the invention.

Another factor to consider is the following: ee

8 BE2022/5412 b) a density, direction and speed, in particular directional density and

Speed distribution of road users in the side streets along the

Route, in particular the side streets that have not yet been passed in the direction of travel to an upcoming traffic junction, in particular an upcoming intersection, based on the

Real-time data.

Constantly taking all real-time data into account is too computationally intensive and therefore virtually impossible. The invention must therefore make an appropriate selection. The variant according to the invention, factor b), has proven to be particularly suitable and can be carried out using contemporary computing resources, in particular on-the-fly. In one of our embodiments, tuples come from vehicle speeds and

Street side/direction of travel for use.

Simulations have shown that the simultaneous consideration of aspects a) and b) has a better effect on the overall dynamics of traffic than a purely additive approach

Summing up the effects would suggest. Synergy effects from factors a) and b) must therefore be present, since an effect[,a) and b)"] according to the invention exceeds the sum of the individual effects, effect[,a)"] + effect["b)"]. .

For example, if 10% of the road users have taken factor a) into account and another 10% have taken factor b) into account (the remaining 80% are conventional behavior), but the two sets of road users are deliberately chosen to be disjoint, a measurable improvement is already noticed in the simulation (Scenario A). However, this result is worse than what occurs when the 10% coincide. If 10% of road users take factors a) and b} into account (the remaining 90% use the conventional method) (scenario B), the result of the overall traffic dynamics is better than in scenario A. In

Essentially, this means that the average road user can get to the road faster

The goal is what saves time and resources and is suitable for making a small contribution to environmental protection, which, however, increases when the invention is used on a large scale

9 BE2022/5412 certainly “laps”. In numerical terms, the smallest improvements ultimately have an enormous reach if they are scaled accordingly.

In the range of approx. 65 - 80% of road users who use Effect[,a)"] + Effect[,b)"] according to the invention and are trying to

When the speed recommendation (autonomous, semi-autonomous or manual) is driven, a kind of slow saturation effect occurs on the traffic dynamics. In particular, it is not necessary that all road users without exception use the invention in order to achieve the advantageous effects. In particular, a small remaining amount of conventional road users and obstacles does not detract from the beneficial effects.

This “autonomy” is a special achievement of the invention, which allows it to differentiate itself from conventional coordination-based methods.

If a revision condition occurs, the speed recommendation is revised and the best-rated route is also taken into account again

Route suggestions checked.

Revision conditions can be of a wide variety of nature. Especially in the ; In different fields of application, a wide variety of revision conditions can be implemented within the scope of the invention.

An example revision condition is driving through an intersection or an

transport hub.

The dependent claims create several different revision conditions which have proven to be particularly suitable. Reference is therefore made to the discussion at the appropriate point.

EEE

10 BE2022/5412

According to further training, at least one other road user will...

Real-time data and/or historical data, takes into account its own start and

Destination point is unknown for the purposes of the route finding method.

This feature can be viewed as a permissible “disclaimer” feature. If the starting and finishing points of all participants in a theoretical system are known, there are further options for specifically evaluating and using this information.

However, an enormous advantage of the present invention is that this is the case

Practice is given less often - information situation not to be dependent, and yet

Deliver results with excellent asymptotic behavior.

The invention provides a type of “effective group intelligence” based on the combination of individual operational intelligences.

According to further training, at least one other road user will be with the

Finding the route takes into account its own start and destination points for the purposes of

Method for finding a route is also known, and a game theory method is also used to generate route suggestions and / or route revisions, which provide a synergistic, coordinated behavior of the

Enable road users.

However, if such data, in particular the starting point and desired destination, are known, for example for a subset of road users, the application of a game-theoretic method to this subgroup of road users can be further

Generate improvement. The expert knows suitable cooperative and non-cooperative models. Cooperative models are for the majority of people

Applications/uses of the invention are particularly suitable because in a variety of

Use cases either someone operates an overall logistics system (i.e. in the

Usually a single party, for example in warehouse logistics) or it, as in — | Ee

11 BE2022/5412

Road traffic, a certain level of consideration is important anyway. Non-cooperative models can also be used, especially if the models used are cooperative due to so-called self-enforcement

Deliver behavioral results. The result is that the behavior is determined by an “extended

S “group intelligence”, which specifically introduces coordinated behavior in addition to the individual-oriented basic framework.

According to further training, a revision condition occurs when the actual

Speed of a vehicle over a specific time interval, in particular a contiguous time interval, deviates from the speed recommendation by at least a specific speed difference, in particular an absolute and/or relative speed difference.

Since failure to follow the speed recommendation to a large extent means that one's own route can no longer be followed according to the forecast and also has a negative influence on the other road users and possibly their routes, the revision under this revision condition leads to an improvement the

Overall dynamics of a transport system. It may be that from the

Speed recommendation by a manual driver based on a conscious

decision was deviated from. But it can also occur in the context of autonomous or semi-autonomous driving.

According to a further development, data on traffic light switching in the traffic system is taken into account, in particular: real-time data on traffic light switching, which is provided directly by a traffic light control, and / or real-time data

Traffic light switching, which is recorded by suitable sensors and/or

12 BE2022/5412 historical data, in particular statistical data, on the duration and frequency of certain phases of a traffic light switching.

Taking this data into account allows the behavior to be excellently adapted to the traffic light circuits (or general “traffic sign systems”). The coordinated behavior improves traffic dynamics. There are synergistic ones

Effects with the (other) real-time data, for example from terrestrial sensors. The simultaneous

Tapping into traffic light switching behavior and using terrestrial sensors for

Real-time data on road users leads to the predictions/evaluations of the

Artificial intelligence can be further improved.

According to a further development, the method further comprises a step of sending a request to switch a traffic light according to the request, in particular switching to

Green.

Although this is not currently the current reality, at least in reality

Road traffic systems - i.e. the StVO - corresponds, the simulations have shown that a further, albeit small, contribution to traffic dynamics by allowing this

option can be achieved.

This is particularly the case at less busy intersections/traffic signs. Therefore, a classification into two groups is particularly useful: traffic lights that can be switched by participants and traffic lights that cannot be switched. Alternatively, the latter traffic lights can only be switched when the

The intersection is exceptionally poorly attended, which can be the case, for example, in the early afternoon or at night. a.

13 BE2022/5412

According to a further development, the step of sending the request takes place in response to the occurrence of a revision condition. This improves the dynamics. A route that was originally adapted to the existing switching behavior of the traffic sign systems using artificial intelligence should be revised, or at least re-examined, if one

Revision condition occurs. It is often the case that the switching behavior of the traffic lights, especially the next traffic light, is no longer beneficial for the road user. It therefore has to be “improvised”. Switching the traffic lights, if approved, represents an efficient solution as long as this option is available at a critical time. In particular, this avoids chain reactions in the transport system, such as “domino effects”, in which, for example, the deviation from the plan of one road user affects the route plans of numerous other road users, all of which would then have to be revised, for example.

According to a training course, a revision condition occurs if the request is not granted and/or the request remains unanswered for a predetermined period of time.

The other way around, it has also been shown that the initiation and implementation of a revision of the route under this condition has an overall positive effect on the key performance of the route

impact the transport system. In particular, requests that have not been answered will be viewed as “no”. According to a further development, however, a user (or, for example, a

Technical Support/Maintenance) will be informed about such unanswered inquiries so that appropriate measures can be taken so that future inquiries can be answered reliably and with a quick response time.

However, according to further training, the route does not have to be revised if one

The revision condition had already occurred. This can be particularly the case if the request itself was sent due to a revision condition.

14 BE2022/5412

According to a further development, a revision condition occurs when, according to a simulation, a right-of-way phase, in particular green phase and / or phase formed from green and

Yellow phase, a signaling system, in particular a traffic light, at an upcoming traffic junction, in particular an upcoming intersection, can no longer be reached in compliance with a traffic rule. This revision creates a more suitable route for the road user themselves, while at the same time avoiding the negative effects on the overall transport system that would occur if the road user were to maintain their route without being revised.

According to a further development, the first artificial intelligence takes into account that the route of another road user who enjoys and/or claims priority for a specific route and authenticates himself by means of a certificate, according to a predetermined and/or transmitted priority factor in order to avoid the route by the road user is devalued in favor of the other road user who enjoys and/or claims priority. This feature is particularly important to use for, for example, the police, fire brigade and especially emergency doctors

Ambulance. However, in another example, this can also be done for VIPs like the

Austrian Federal Chancellor.

According to a further development, real-time data is also generated by flying measurement and/or

Sensor devices, in particular flight drones, are provided. Such data can in particular complement terrestrial sensor data. In some cases and/or in some areas they may replace the latter. The drones can be used flexibly and also have a large catchment area. The acquisition costs are therefore lower.

15 BE2022/5412

However, the maintenance/flight costs can be similarly high or even higher because

Flying objects have to be kept in the air permanently using energy.

In a preferred embodiment, flying drones are used in the urban agglomeration area of a city. While terrestrial sensor technology is used in the city center, flying sensors are used in the suburbs/agglomeration of the respective city

Measuring and/or sensor devices are used. This contributes to the less dense

settlement and the more manageable traffic situation in these areas

Invoice.

Training of the entire system of muttipler artificial intelligences takes place, for example, through the following steps: Training the first artificial intelligence with an assigned data set on routes, actual time courses of a past one

Covering these routes and success factors, especially through a monitored one

Learning and/or training the second artificial intelligence to find a best-rated route through real-time data and historical data, in particular through unsupervised learning. The control loop created by the two artificial intelligences according to the invention is particularly stable.

Self-supervised learning can also be used within the scope of the invention.

The invention creates relevant methods, including methods for training one or both classes, computers, computer programs, computer networks, data and computer-readable data carriers.

16 BE2022/5412

All features disclosed in connection with a claim category, for example features disclosed in the context of the corresponding procedure, can be included in the

Connection with the other claim categories, for example devices and

Computer programs can be used, and vice versa. The

Claim categories do not impose any categorical restrictions in this regard

Invention features. The same applies to several independent claims within the same

Claim category, if any.

The invention can be used in numerous fields of application/technical applications

be brought into use. The route generation and/or route revision for the

Road traffic is just one example. Navigation systems for vehicles, autonomous

Driving, semi-autonomous driving and driver assistance systems can also benefit from the invention.

The following additional areas can be provided: Navigation of flying objects such as

Drones, aircraft, flying objects, delivery robots, e-scooters, air cargo, integrated circuits,

Bus systems in system and/or electrical engineering, suspension railways, self-propelled ones

Vehicle transport systems, in particular self-propelled tunnels, department store

Shopping assistants, department store logistics, especially automatic restocking and/or

Cobots, cleaning robots, luggage transport robots.

In department store logistics, for example, there is also the problem that...

Block road users from each other. In many warehouses and/or warehouses, it is not the total space that is the critical resource (the “bottle neck”), but rather the time-efficient one

Access to the individual product boxes.

For example, the invention relates to logistics robots (such as the group of

Types “Amazon Kiva”).

17 BE2022/5412

In another example, real-time data is recorded by drones, possibly also about vehicles on the ground, but in particular also about the drones themselves. The drone flight of the drones that collect real-time data can also be monitored by the drones themselves

Benefit from the advantages of the invention. The drone flight is particularly efficient, especially with regard to the overall system and the overall dynamics achieved by the drone flight.

If these drones can be used to obtain real-time data on other objects such as

Vehicles on the ground are recorded, ground-based traffic systems - already based on the invention and its use in drones - can benefit from more available and more frequently updated real-time data, while they themselves in turn

Use invention. So the use of the invention itself can be used again.

This creates a meta level. It is also possible to create several meta-levels or multi-stage meta-levels of this type for using the invention.

Essentially, the invention can be used in all “traffic systems” in the broadest sense, where several road users could block each other and/or share a resource (e.g. space/time/the like).

For example, parts of a computer system often share a so-called

Communication bus. Only one (or a limited number of) participants can be connected at any given time via a communication bus (in a very simple

Example) communicate. This can also be used in integrated circuits.

In all systems there is always a priority question and therefore, in the broadest sense, a route question.

Even when shopping in the supermarket, customers regularly have problems “driving” their shopping carts. The invention can therefore also be used in Assisted

18 BE2022/5412

Shopping can be used. For example, customers' shopping carts or other suitable vehicles are no longer used manually, uncontrolled and with restrictions

Information somehow brought to the desired product, but it is from

Customers only select a product (e.g. via a display or another

Human User Interface), and the vehicle is brought to the desired destination using the present invention.

The invention can also be used in agriculture. For example, work systems with comprehensive robots are used here. These robots can take on a wide variety of tasks, for example Precision

Spraying of herbicides possible. These robots currently usually drive on a ground-based basis, but flying work robots are also conceivable.

The invention can also be used in rail systems. For example, these can be automatic storage and retrieval systems (e.g. Auto-Store). Here, for example, routes are taken both by robots on rail systems and also by lifts with which these robots are equipped. With the invention, both aspects can be taken into account and a more productive dynamic is achieved, taking both the rail system and the lift system into account.

Classic rail systems (such as trains on railway tracks, for example with switches) are also possible.

Robot systems that orient themselves along lines (for example, travel along rails) can also be optimized by the invention.

Edge computing and/or cloud computing can be used.

The invention can be used in the field of process optimization, including

Business process optimization is used. There are also defensive purposes

19 BE2022/5412 conceivable. All other technical fields that benefit from faster movement are conceivable as fields of application.

Further system variants that can benefit from the present invention are, for example, the following:

Sorting robots in department stores, for example, supply conveyor belts and can implement incoming and/or outgoing goods. Order too

Processing is possible, for example.

Automated Guided Vehicles (AGVs), for example, are driverless transport vehicles. which can, for example, pull/carry loads or transport material.

In addition to more classic road traffic participants (automobiles), there are also more modern participants, such as e-scooters.

Nanorobots can be used, for example, in the medical sector. These can, for example, carry out medical repair work.

Microrobots can take on defense and/or observation tasks. For example, the microrobots are equipped with microweapons.

The numerous examples explain the many possible options in an exemplary manner

Areas of application for the use of the present invention. However, the invention in no way wants to be limited to the examples mentioned alone. The use of the invention is also possible and intended for other systems not mentioned here.

Internet traffic can also benefit, as it is usually broken down into smaller units (IP packets), all of which can take different routes through a traffic-intensive (and fundamentally bandwidth-limited) network. A special application arises in the area of secret sharing technology (sharing secrets using technical means).

20 BE2022/5412

The invention can therefore be used for route optimization with time savings for a wide variety of applications and systems, and on a wide variety of size scales. In principle, this can only be achieved through appropriate access to the

System (including for data reference), for example, software according to the invention can be used in order to be able to implement and use route optimization in existing systems. This is a big advantage because the invention is extremely system and platform independent and can be transferred to a wide variety of technical systems.

Although some aspects have been described in the context of a device, it is clear that these aspects also represent a description of the corresponding method, where a block or a device corresponds to a method step or a function of a

Procedural step corresponds. Analogous to this, aspects that arise within the framework of a

Process step are described, also a description of a corresponding one

Blocks or elements or a property of a corresponding device.

Embodiments of the invention can be implemented in a computer system.

The computer system may be a local computing device (e.g., personal computer,

laptop, tablet computer or mobile phone) with one or more processors and one or more storage devices or may be a distributed computing system (e.g. a cloud computing system with one or more processors or one or more storage devices distributed in different locations, for example). Example of a local client and/or one or more remote server farms and/or

data centers). The computer system may be any circuit or combination of

Include circuits. In one embodiment, the computer system may have one

Se

21 BE2022/5412 or more processors, which may be of any type. According to local

By use, processor can mean any type of computing circuit, such as, but not limited to, a microprocessor, a microcontroller, a

Complex instruction set microprocessor (CISC), a reduced-level microprocessor

Instruction Set (RISC), a Very Long Instruction Word (VLIW)

microprocessor, a graphics processor, a digital signal processor (DSP), a multi-core

Processor, a field programmable gate array (FPGA), or any other type of

Processor or processing circuit. Other types of circuits included in the

Computer system may include a custom-built circuit, an application-specific integrated circuit (ASIC), or the like, such as one or more circuits (e.g., a communications circuit) for use in wireless devices such as. B. mobile phones, tablet computers, laptop computers,

radios and similar electronic systems. The computer system may include one or more storage devices that include one or more

May include memory elements that are suitable for the respective application, such as a main memory in the form of a random access memory (RAM,

Random Access Memory), one or more hard drives and/or one or more

Drives that handle removable media such as CDs, flash memory cards, DVDs and the like. The computer system may also include a display device, one or more speakers, and a keyboard and/or controller that includes a

Mouse, trackball, touch screen, voice recognition device or any other

Device may include that allows a system user to enter information into the

to enter and receive information from the computer system.

Some or all of the method steps can be carried out by (or using) one

Hardware device is executed, such as a processor

Be a microprocessor, a programmable computer or an electronic circuit

22 BE2022/5412 can. In some embodiments, one or more of the most important

Process steps are carried out by such a device.

Depending on particular implementation requirements, embodiments of the invention may be implemented in hardware or software. The implementation may be with a non-volatile storage medium such as a digital storage medium such as a floppy disk, a DVD, a Blu-Ray, a CD, a ROM, a

PROM and EPROM, an EEPROM or a FLASH memory, on which electronically readable control signals are stored, which interact (or can interact) with a programmable computer system so that the respective method is carried out. Therefore, the digital storage medium can be computer readable.

Some embodiments according to the invention include a data carrier with electronically readable control signals that can interact with a programmable computer system to perform one of the methods described herein.

In general, embodiments of the present invention may be as follows

Computer program product can be implemented with a program code, whereby the

Program code for the execution of one of the methods is effective if the

Computer program product running on a computer. The program code can, for example, be stored on a machine-readable carrier.

Further exemplary embodiments include the computer program for carrying out one of the methods described herein, which is stored on a machine-readable medium.

23 BE2022/5412

In other words, an embodiment of the present invention is therefore a

Computer program with a program code for carrying out one of the methods described herein when the computer program runs on a computer.

A further embodiment of the present invention is therefore a storage medium (or a data carrier or a computer-readable medium) which has a data stored thereon

Computer program for carrying out any of the methods described herein when executed by a processor. The data carrier, the digital one

Storage medium or the recorded medium is usually tangible and/or not seamless. Another embodiment of the present invention is one

Device as described herein, comprising a processor and the storage medium.

A further exemplary embodiment of the invention is therefore a data stream or a

Signal sequence that enables the computer program to carry out one of the described herein

represents procedure. The data stream or the signal sequence can, for example, be configured so that it is transmitted via a data communication connection, for example via the

Internet, are transmitted.

Another embodiment includes a processing means, for example a

Computer or programmable logic device configured or adapted to perform any of the methods described herein.

Another exemplary embodiment includes a computer on which the

Computer program is installed to perform any of the methods described herein.

A further exemplary embodiment according to the invention comprises a device or a

System that is configured to (e.g. electronic or optical) a

Computer program for carrying out one of the methods described herein

to transmit to the recipient. The recipient can be, for example, a computer, a mobile

Device, a storage device or the like. The device or that

24 BE2022/5412

For example, the system may include a file server for transmitting the computer program to the recipient.

In some embodiments, a programmable logic device (e.g., a field programmable gate array, FPGA) may be used to control some or all

To carry out the functionalities of the methods described herein. In some

Embodiments can have a field-programmable gate arrangement with a

Microprocessor cooperate to perform any of the methods described herein. In general, the procedures are preferably carried out by everyone

Hardware device performed.

Embodiments may be based on using a machine learning model or

Machine learning algorithm based. Machine learning can refer to algorithms and statistical models that computer systems can use to perform a specific task without using explicit instructions, rather than relying on

Leaving models and inference. For example, in machine learning, instead of a rule-based transformation of data, a transformation of data can be used that can be derived from an analysis of historical and/or training data. For example, the content of images can be changed using a

Machine learning model or using a machine learning algorithm. In order for the machine learning model to analyze the content of an image, the machine learning model can use training images as input and

Training content information can be trained as output. By training the

Machine learning model with a large number of training images and/or

Training sequences (e.g. words or sentences) and associated ones

Training content information (e.g. labels or notes) “learns” that

Machine learning model to recognize the content of the images, so that the content of images that are not included in the training data using the machine learning model

25 BE2022/5412 can be recognized. The same principle can be used for other types of sensor data as well: By training a machine learning model using

Training sensor data and a desired output, the machine learning model "learns" a transformation between the sensor data and the output, which can be used to generate an output based on non-learning models provided to the machine learning model.

Provide training sensor data. The data provided (e.g. sensor data,

Metadata and/or image data) can be preprocessed to obtain a feature vector, which is used as input for the machine learning model.

Machine learning models can be trained using training input data. The examples above use a training method called ,Supervised

“Learning” is called. In supervised learning, the machine learning model is used

trained using a plurality of training samples, each sample being one

A plurality of input data values and a plurality of desired output values, i.e. H. Each training sample is assigned a desired output value, which can include.

By specifying both training samples and desired output values, the machine learning model “learns” what output value should be based on one

Input sample that is similar to the samples provided during training is to be provided. In addition to supervised learning, semi-supervised learning can also be used

Learning can be used. In semi-supervised learning, some of the training samples lack a desired output value. Supervised learning can be done on a

Supervised learning algorithm based (e.g. a classification algorithm, a

regression algorithm or a similarity learning algorithm).

Classification algorithms can be used when the outputs are limited to a limited set of values (categorical variables), i.e. H. the input is classified as one of the limited set of values. Regression algorithms can be used when the outputs are of any numerical value (within a range)

26 BE2022/5412. Similarity learning algorithms can perform both classification and

Regression algorithms may be similar, but are based on learning from examples

Using a similarity function that measures how similar or related two objects are. In addition to supervised learning or semi-supervised learning, unsupervised

Learning can be used to train the machine learning model. At the

Unsupervised learning may (only) provide input data and a

Unsupervised learning algorithm can be used to create a structure in the

Find input data (e.g. by grouping or clustering the input data, finding commonalities in the data). Clustering is the allocation of input data that a

A plurality of input values are divided into subsets (clusters), so that input values within the same cluster are divided according to one or more (predefined)

Similarity criteria are similar while dissimilar to input values included in other clusters.

Reinforcement learning is a third group of machine learning algorithms. In other words, reinforcement learning can be used to train the machine learning model. Reinforcement learning involves training one or more software actors (so-called “software agents”) to perform actions in an environment.

Based on the actions taken, a reward is calculated.

Reinforcement learning is based on training one or more software

agents to select the actions in such a way that the cumulative reward is increased, resulting in software agents that become better at the task they are given (as evidenced by increasing rewards).

Furthermore, some techniques can be applied to some of the machine learning algorithms. For example, feature leaching can be used. In other words, this

Machine learning model may be at least partially trained using feature learning, and/or the machine learning algorithm may be a feature learning

27 BE2022/5412

Include component. Feature Learning Algorithms, Representation Learning

Algorithms can preserve the information in their input but transform it in such a way that it becomes useful, often as a pre-processing step before the

Perform classification or prediction. For example, feature learning can be based on principal component analysis or cluster analysis.

In some examples, anomaly detection (i.e., outlier detection) may be used, which aims to provide identification of input values that

raise suspicion because they differ significantly from the majority of input and training data. In other words, the machine learning model can be trained at least in part using anomaly detection and/or machine learning

Algorithm may include an anomaly detection component.

In some examples, the machine learning algorithm may take a decision tree as

Use predictive model. In other words, the machine learning model can be based on a decision tree. With a decision tree you can

Observations about an object (e.g. a set of input values) by the

Branches of the decision tree are represented, and an output value corresponding to the

The object corresponds to can be represented by the leaves of the decision tree.

Decision trees can have both discrete values and continuous values

Support output values, When discrete values are used, the decision tree can be called a classification tree, when continuous values are used, the decision tree can be called a regression tree.

Association rules are another technique that can be used in machine learning algorithms. In other words, the machine learning model can be based on one or more

Association rules based. Association rules are created by identifying relationships between variables in large sets of data. The machine learning

Algorithm may identify and/or utilize one or more ratio rules that

28 BE2022/5412

Represent knowledge that is derived from the data. The rules can e.g. B. be used to store, manipulate or apply knowledge.

Machine learning algorithms are usually based on a machine learning model.

In other words, the term “machine learning algorithm” can mean a set of instructions that can be used to create, train, or use a machine learning model. The term “machine learning model” can be one

data structure and/or a set of rules representing the learned knowledge (e.g. based on that executed by the machine learning algorithm

Training). In embodiments, the use of a machine learning

Algorithm the use of an underlying machine learning model (or a

majority of underlying machine learning models). The use of a machine learning model may imply that the machine learning model and/or the

Data structure/set of rules which is/are the machine learning model by a

Machine learning algorithm is trained.

For example, the machine learning model can be an artificial neural network (ANN). ANNs are systems inspired by biological neural networks such as those found in a retina or brain. ANNs include a plurality of interconnected nodes and a plurality of connections, called edges, between the nodes. There are usually three types of nodes,

Input nodes that receive input values, hidden nodes that (only) communicate with others

Nodes are connected, and output nodes that provide output values. Each node can represent an artificial neuron. Each edge can send information from one

node to the other. The output of a node can be expressed as a (nonlinear) function of the

Inputs must be defined (e.g. the sum of its inputs). A node's inputs can be entered into the function based on a "weight" of the edge or node that is the

provides input. The weight of nodes and/or edges can be 1

29 BE2022/5412 can be adapted in the learning process. In other words, training an artificial neural network may include adjusting the weights of the nodes and/or edges of the artificial neural network, i.e. H. to achieve a desired output for a given input.

Alternatively, the machine learning model can be a support vector machine, a random

Forest model or a gradient boosting model. Support Vector Machines (i.e.

Support vector networks) are supervised learning models with associated learning algorithms that can be used to analyze data (e.g. in a classification or

regression analysis). Support vector machines can be trained by providing input with a plurality of training input values belonging to one of two categories. The Support Vector Machine can be trained to be either

Assign categories a new input value. Alternatively, the machine learning

Model be a Bayesian network, which is a probabilistic directed acyclic graphical model. A Bayesian network can represent a set of random variables and their conditional dependencies using a directed acyclic graph. Alternatively, the machine learning model may be based on a genetic algorithm, which is a search algorithm and heuristic technique that mimics the process of natural selection.

N

30 BE2022/5412

Figure list

The present invention is explained below with reference to the schematic figures

Drawings specified exemplary embodiments explained in more detail. It shows:

1 shows a schematic representation of the invention according to an embodiment,

Fig. 2 is a schematic representation of an exemplary map section

road transport system,

Fig. 3 is a schematic representation of a deep neural network for use as

Artificial intelligence within the scope of the present invention.

In all figures, identical or functionally identical elements and devices - unless otherwise stated - have been given the same reference numerals.

; 31 BE2022/5412

Description of the drawings

Figure 1 shows a schematic representation of the invention according to one

Embodiment.

The real-time data 101 and the historical data 102 serve as input for the first

Artificial Intelligence 200. A top-rated route 103 can be made from a variety of

Route suggestions can be selected. The second artificial intelligence 300 for

In addition to this route, the success assessment also has the spatiotemporal course 104 of the route actually traveled by a road user available. A

Control loop is created by the feedback of this data into the historical data 102.

A symbol was placed in the middle of the feedback loop 301 just for clarity

Clarification of this control relationship, comprising the two artificial intelligences 200, 300, arranged (reference number 301).

An artificial intelligence in the sense of the invention is any machine learning model, but in particular also deep learning models.

A simulation 105 of spatiotemporal dynamics serves as input for the

Speed recommendation 106.

When a revision condition 107a occurs, a revision 107b is triggered. This applies in particular to the best-rated route 103 as well as the current ideal speed 106.

This system optimizes traffic and allows participants to reach their destinations more quickly

Goals. This affects the individual participants who use the invention as well as the holistic dynamics of the transport system.

32 BE2022/5412

Figure 2 shows a schematic representation of an exemplary map section of a road traffic system. In the road traffic system there are vehicles 30 and terrestrial sensors 40 along the roads. Five routes or route suggestions 1, 2, 3, 4, 5 between the starting point 10 chosen by/for a road user and a corresponding destination point 20 are shown. With the invention, for example, route suggestions 1, 2, 3, 4 and 5 are each evaluated. A best-rated route is then selected. The revision conditions include, among other things, new ones

Examinations initiated. ;

Figure 3 shows a schematic representation of a deep neural network for use as a machine learning model in the context of the present invention. For example, such a network is used in the first and/or second artificial intelligence. The

Deep Neural Network (Artificial Neural Network) has at least one

Input neuron 301, usually several, especially numerous (the designation 301a-z is merely symbolic and in no way limits a number). There is at least one

Output neuron 303. Various structures can be arranged between these, in particular one or more hidden layers 302.

EN En a

33 BE2022/5412

Reference symbol list 1 Route 1/Route suggestion 1 2 Route 2/Route suggestion 2 3 Route 3/Route suggestion 3 4 Route 4/Route suggestion 4 5 Route 5/Route suggestion 5

Starting point 10 20 Destination point 30 Road user/vehicle 40 Terrestrial sensor 100a Device according to the invention according to an embodiment 101 Real-time data 102 Historical data 103 Best rated route 104 Actual time course of a route traveled 105 Spatio-temporal system simulation 106 Currently recommended ideal speed 107a Revision condition 107b Revision 200 First AI 200a Example map ography of a road transport system 300 Second AI 300a Artificial neural network 301 Control/control circuit/feedback loop 301-(a-z) exemplary input neurons 302 exemplary hidden layers (schematic) 303 exemplary output neurons

Claims (16)

REG > ot SY Al 34 BE2022/5412 patent claims 1. Method for finding a route for a road user (30) in a traffic system, in particular the fastest possible route with a short traversion time, in particular a short travel time, for the road user through the traffic system, from a starting point (10) in the system to a destination point (20 ) in the system with road users (30) using big data, comprising: - Providing real-time data (101) in relation to other road users (30) who are in the traffic system, in particular position and speed data, in particular data from terrestrial, local distributed sensor technology (40) - providing historical data (102) on routes (1-5) traveled by road users in the transport system in the past, together with success factors assigned to them, which quantify a comparison of the actual routes traveled with a respective original prediction and using a second artificial intelligence method (300) - providing several route suggestions (1-5) as a preselection for a route between the starting and destination points for the road user and evaluating these route suggestions (1-5) by a first artificial intelligence method ( 200) based on the real-time data (101) and the historical data (102) including success factors, whereby a temporal development of the road users over the course of the route is at least partially simulated (105) - Selecting the best-rated route (1-5) a | 1 35 BE2022/5412 - Calculating an instantaneous ideal speed (106) as ; Speed recommendation for driving along the route, in particular based on a) a current and/or expected traffic light switching at an upcoming traffic junction, in particular an upcoming intersection, and b) a density, direction and speed, in particular direction-related density and speed distribution, of road users in the side streets of the route, in particular the side streets that have not yet been passed in the direction of travel to an upcoming traffic junction, in particular an upcoming intersection, based on the real-time data - revising the speed recommendation (106) and re-checking the route suggestion when a revision condition occurs (107a). 2. The method according to claim 1, wherein at least one other road user is taken into account, through real-time data and / or historical data, whose own starting and destination points are unknown for the purposes of the method for finding a route. 3. The method according to claim 1 or 2, wherein at least one other road user, through real-time data and / or historical data, is taken into account when finding the route, whose own starting and destination points are known for the purposes of the method for finding a route, and Furthermore, a game theory process is used to provide route suggestions and/or 7 36 BE2022/5412 to generate route revisions that enable synergistic, coordinated behavior of road users. 4. Method according to one of the preceding claims, wherein a revision condition occurs when the actual speed of a vehicle over a certain time interval, in particular a continuous time interval, deviates from the speed recommendation by at least a certain speed difference, in particular an absolute and / or relative speed difference. 5. Method according to one of the preceding claims, wherein data on traffic light switching in the traffic system are taken into account, in particular: - real-time data on traffic light switching, which is provided directly by a traffic light control, and / or - real-time data on traffic light switching, which is recorded by suitable sensors and /or - Historical data, in particular statistical data, on the duration and frequency of certain phases of a traffic light switching. 6. The method according to any one of the preceding claims, further comprising a step of sending a request to switch a traffic light according to the request, in particular switching to green. 7. The method of claim 6, wherein the step of sending the request occurs in response to the occurrence of a revision condition. a 37 BE2022/5412 8. The method of claim 6, wherein a revision condition occurs if the request is not granted and/or the request remains unanswered for a predetermined time. 9. The method according to any one of the preceding claims, wherein a revision condition occurs when, according to a simulation, a right-of-way phase, in particular green phase and / or phase formed from green and yellow phase, a signaling system, in particular a traffic light, at an upcoming traffic junction, in particular an upcoming intersection , can no longer be reached while complying with traffic regulations. 10. The method according to any one of the preceding claims, wherein the first artificial intelligence takes into account that the route of another road user who enjoys and/or claims priority for a specific route and authenticates himself by means of a certificate, according to a predetermined and/or transmitted priority factor in order for the road user to avoid the route, it is downgraded in favor of the other road user who enjoys and/or claims priority. 9. Method according to one of the preceding claims, wherein further real-time data is also provided by flying measuring and/or sensor devices, in particular flying drones. 38 BE2022/5412 10. Method for training a first artificial intelligence (200) and/or a second artificial intelligence (300) for the purpose of enabling a method according to one of the preceding claims, comprising the following steps: - Training the first artificial intelligence (100). an assigned data set on routes, actual time courses of a past journey of these routes and success factors, in particular through self-monitored learning and/or supervised learning and/or training of the second artificial intelligence (200) to find a best-rated route through real-time data and historical data, in particular through self-supervised learning and/or unsupervised learning. 11. The method according to claim 10, wherein the first (200) and the second (300) artificial intelligence are arranged relative to each other with respect to the data flow in a control loop (301). 12. Device (100), in particular computer, computer network or computing cloud, which is set up to effect a method according to one of the preceding claims. 13. Computer program which, when executed, effects a method according to one of claims 1-11. 39 BE2022/5412 14. Route data which makes a route (1-5) identifiable which was determined and/or revised according to a method according to one of claims 1-11. 15. Computer-readable storage medium comprising the computer program according to claim 13 and / or the route data according to claim 14. 16. Use of a method according to one of claims 1 - 11, a device according to claim 12, a computer program according to claim 13, route data according to claim 14 or a computer-readable storage medium according to claim 15 for route generation and / or route revision in the following areas: navigation systems for Vehicles, autonomous driving, semi-autonomous driving, driver assistance systems, smartphone applications or online services, drones, aircraft, flying objects, delivery robots, e-scooters, air cargo, integrated circuits, bus systems in system and/or electrical engineering, monorails, self-driving vehicle transport systems , in particular self-propelled tunnels, department store shopping assistants, department store logistics, in particular automatic restocking and/or cobots, cleaning robots, luggage transport robots, robot work systems, in particular in agriculture, car store systems, sorting robots, in particular in department stores and warehouses, order processing, in particular for incoming and/or outgoing goods, nanorobots, in particular medical nanorobots, defense robots, in particular micro-defense robots, in particular micro-defense robots equipped with microweapons, sending data packets in networks, for example the Internet, as well as other areas in which a transmission bandwidth, 40 BE2022/5412, particularly locally, is limited, secret sharing techniques, in particular secret sharing techniques.