EP4010767A1

EP4010767A1 - Communication with automatable industrial devices or systems or with the controller thereof

Info

Publication number: EP4010767A1
Application number: EP20749875.9A
Authority: EP
Inventors: Slawomir Sander; Christopher Schwarzer
Original assignee: KUKA Deutschland GmbH
Current assignee: KUKA Deutschland GmbH
Priority date: 2019-08-07
Filing date: 2020-07-30
Publication date: 2022-06-15
Also published as: CN114556880A; WO2021023614A1; US20220407797A1; US11799759B2; CN114556880B; DE102019211843A1

Abstract

The invention relates to a method for establishing a communication connection between a first node (11) in a first network of at least two networks and a second node (31) in a second network of the at least two networks. At least one node of the first and second nodes (11, 31) is an automatable industrial device (e.g. a robot) or an automatable industrial system or a controller for same. Each of the at least two networks individually forms a homogeneous address space, but the at least two networks do not form a homogeneous address space together. The method has at least the following steps: transmitting (101) a call to establish the communication connection between the first node and the second node, said call having routing path information (10/20/30/31) which indicates the routing path from the first node (11) to the second node (31), wherein the routing path information (10/20/30/31) has at least one identifier (10, 20, 30, 31) for each of the networks or nodes to be traversed on the routing path but not necessarily an identifier for the first network; and establishing (102) the communication connection between the first node (11) and the second node (31) on the basis of the routing path information (10/20/30/31). The invention additionally relates to a corresponding control device, a computer program, and a computer program product.

Description

description

Communication with automatable industrial devices or systems or with their control

The present invention relates to communication, in particular setting up a communication link, with automatable industrial devices or systems or with their control.

Automated industrial devices include in particular autonomous transport devices (automated guided vehicles, AGV), automatable processing devices or stations, stationary or mobile manipulators, stationary or mobile manipulator arrangements, in particular stationary or mobile robots or stationary or mobile robot arrangements, or controls therefor, or combinations of two or more of these. Automatable industrial systems include, in particular, arrangements that have one or more automatable industrial devices, in particular one of the devices mentioned above. A controller that is considered here includes controllers that are used to control the above-mentioned devices or systems. These can be designed, for example, as a computer, computer system, control card or integrated circuit.

It is known to communicate with an automatable industrial device or system or with its control, for example via a GRPC interface via the Protobuf protocol, in particular to address such a device, system or control or to exchange data with it. In order to be able to address GRPC services across network borders, at least one VPN network (Virtual Private Network) is necessary. In addition, every component within the VPN must be clearly addressable.

The object of the present invention is to improve the communication or the establishment of a communication link with an automatable industrial device or installation or its control. This object is achieved by a method with the features of claim 1. Claim 14 provides protection for a control device which is set up to carry out a method described here. Claims 15 and 16 provide a computer program or a computer program product for performing a method described here under protection. The subclaims relate to advantageous developments.

A first aspect of the present invention relates to a method for setting up a communication connection between a first node in a first of at least two networks and a second node in a second of the at least two networks, at least one of the first and second nodes being an automatable industrial device or is an automatable industrial plant or a controller for this, the at least two networks individually each forming a homogeneous address space, the at least two networks together not forming a homogeneous address space, the method having the following steps:

Sending a call to set up the communication link between the first node and the second node, the call having routing path information indicating the routing path from the first node to the second node, the routing path information at least one identifier of each of the Have routing path to networks or nodes to be traversed, but not necessarily have an identifier of the first network; and

Establishing the communication link between the first node and the second node based on the routing path information.

In the present case, a communication connection is understood in particular to be a connection that enables data exchange in at least one direction, in particular in two directions, for example for the purpose of one Controlling an automatable industrial device or system, or to determine measured values in relation to this. It should be noted here that the control of the device or system does not necessarily have to take place immediately, but can be saved for later use.

A network with a homogeneous address space is, for example, the Internet, a cloud network, a local network or an individual device such as a computer, laptop, smartphone, etc. If you consider each of these networks individually, each node (each component ) of this network can be clearly addressed within the respective network, hence the term homogeneous address space. However, if you consider these networks together as an overall network, they generally do not form a homogeneous address space. H. two nodes can be uniquely addressable within their respective networks by the same identifier (for example a number), but not within the combination of their networks.

In principle, this could be solved by giving each node a globally unique address, or by addressing translation. The inventors of the present invention see both as restrictive and / or inflexible and / or associated with great effort. The method according to the invention provides an alternative approach which is in particular more flexible and / or associated with fewer restrictions and / or costs.

The routing path information uses appropriate identifiers to indicate which networks or nodes are to be traversed on the routing path. On the basis of this routing path information, the communication connection can therefore be set up in particular in a unique manner between the first node and the second node.

It is provided that a call typically originates from the first node, and for this reason the routing path information does not necessarily have to include an identifier of the first network, because this is known from the outset, so to speak. However, the routing path information can also have an identifier of the first network. This can be particularly advantageous if the call is from a higher-level, for example Control device goes out, but it is intended that a communication connection between the first node and the second node is set up within the scope of this call (ie not between the higher-level control device and one of the nodes).

In order to route the call correctly or to determine the routing path correctly, it may be sufficient if the routing path information only contains the identifiers of the networks to be traversed, without alternatively or additionally containing identifiers of the nodes to be traversed in these networks. At the same time, however, it is possible to additionally or alternatively specify the identifiers of the nodes to be traversed in the routing information, in particular if the routing path within a specific network leads via two or more nodes of this network. Even in such a case, however, this is not mandatory because a decision can also be made within a network via which nodes within this network the call is to be routed without this having to be specified in the routing path information of the call.

In one embodiment, the routing path information has an identifier of the second node.

This can be of particular advantage if the network in which the second node is located has several nodes. The call can then be reliably routed to the second node. If the second network has only one node, namely the second node, the second node is in a certain way identical to the second network, so that it is not necessary to specify an identifier for the second node in addition to an identifier for the second network. Likewise, the specification of the identifier of the second node in the routing path information may not be necessary if it is clear from the outset, for example due to the nature of the call, to which node within the second network it is to be routed.

After one execution, the call also has a connection identifier. Such a connection identifier can be used to identify the communication connection, in particular to connect a specific communication connection from another communication connection distinguish. The connection identifier can be assigned in a manner to be described below.

According to one embodiment, the communication connection between the first node and the second node has at least two messages, all messages of this communication connection between the first node and the second node having this connection identifier.

By using the same connection identifier for several messages that belong to a specific communication connection, their association with this specific communication connection can be easily recognized, for example by nodes located on the routing path. Also, the two or more messages can be efficiently grouped in this way. In addition, the use of a connection identifier has advantages to be described for the routing of these messages.

After execution, at least a part of the connection identifier is generated or allocated at random, in particular by the first node. The randomly generated or assigned part of the connection identifier can for example have a number, a letter, a number / letter combination or the like. After this implementation, the connection identifier can be generated or allocated in a relatively uncomplicated manner. Another part of the connection identifier can be generated or allocated in a different way. For example, this part could be generated or allocated based on a current time or an identification of a node, for example the first node. At the same time, however, the entire connection identifier can also be generated or assigned randomly, that is to say not have any part that was not generated or assigned randomly.

The inventors assume that - depending on the implementation - it is relatively unlikely that there are two communication connections that have been set up or are to be set up simultaneously in the overall network and have the same connection identifier. In this respect it is to be expected that it is relatively unlikely that there will be a "collision" between two

There are communication links with the same link identifier. Even this might not be a problem if the corresponding routing paths do not traverse the same networks and / or nodes.

Under certain circumstances, a collision can only be expected if the routing paths for identical connection identifiers lead through the same networks or via the same nodes. Accordingly, it is provided in one embodiment that in the event that the connection identifiers of a first and a second call match, and in particular in the event that the routing paths of the first and second call lead via the same node, the second call is rejected and / or a new connection identifier is generated for the second call or assigned to it.

According to this embodiment, a collision can be avoided or a collision that has already occurred can be “resolved” in a relatively uncomplicated manner. After avoiding or resolving the collision, the risk of a (further) collision can be checked again. This process can be repeated until the collision-free establishment of the communication connections is possible.

After one execution, the associated connection identifier and a node identifier of a predecessor node and a node identifier of a successor node are stored for one or the first call in each node or an associated storage device located on the routing path between the first node and the second node, wherein the predecessor node is a node located on the routing path, in particular a node immediately adjacent on the routing path, in the direction of the first node, and the successor node is a node located on the routing path, in particular a node immediately adjacent on the routing path, in the direction of the second node is.

According to this embodiment, to a certain extent a part of the routing path relevant for this node is stored in each node on the routing path, for example in a routing table. Especially if only the If node identifiers of the neighboring nodes are stored, the memory requirement for this storage is relatively low. In principle, it is not necessary to store non-adjacent node identifiers, but it is possible.

After execution, the routing path information and the associated connection identifier of a call are contained in a start message.

According to this embodiment, there is an association between the routing path information and the connection identifier. These can be transmitted to the appropriate nodes by the start message.

After one execution, messages of this call following the start message have the associated connection identifier but not the routing path information, the messages following the start message preferably based on the nodes on the routing path between the first node and the second node or the associated storage device stored node identifiers are routed.

Because the nodes on the routing path have already been informed of the routing path information in the start message (and the nodes can save the part relevant to them, for example the node identifiers of the neighboring nodes, as explained above, for example in a routing table) it is not necessary that the messages following the start message also contain the routing path information. When a (subsequent) message is received at a node, the node would use the connection identifier to determine the stored node identifier of the neighboring node and forward the message to it.

After execution, when the call is terminated, the stored associated connection identifier and / or the stored node identifier of the previous node and / or the stored node identifier are transferred to the node lying on the routing path between the first node and the second node or the associated storage device. Identifier of the successor node deactivated, in particular deleted. This deactivation or deletion can minimize memory requirements. In addition, this can ensure that information that is no longer required is no longer accessible, which can serve to improve data protection.

According to one embodiment, the communication link is a bidirectional communication link.

In this context, a bidirectional communication link is understood in particular to mean that data can be transmitted, in particular transmitted, in both directions. For example, if the path from the first node to the second node is viewed as the outward path, the path from the second node to the first node represents the return path. In an application example, the first node could be a central robot monitoring device and the second node a robot. On the way there, the first node sends, for example, a request to the second node in order to receive sensor data relating to this second node (robot). The second node determines the sensor data and sends it back to the first node on the way back.

If, as stated above, all nodes on the routing path have an assignment between a connection identifier and the node identifiers of both neighboring nodes, the messages can be routed bidirectionally, i.e. both for the outward and for the return path, based on the connection identifier. Even for a bidirectional communication connection, it is therefore not necessary for messages following a start message to contain routing information relating to the routing path, although this would also be possible as an alternative.

According to one embodiment, the method further comprises: before the call is sent, the first node sends an initial call, the initial call having no or only incomplete routing path information, the information required for the call being based on the initial call Routing path information can be determined, in particular by a node different from the first node, in particular with the aid of an association between information contained in the initial call and the routing path information.

According to this embodiment, the first node does not necessarily have to know the (complete) routing path information that is necessary for setting up the communication link between the first and the second node. Nevertheless, the communication link can be established after this execution.

The initial call could, for example, only identify one type of service that the first node should have provided by another node, the first node not necessarily knowing which other node can provide this service or for providing it Service is provided. On the basis of the initial call or the information contained therein, however, it can be determined, for example by a node different from the first node, which node can or should provide the requested service.

This embodiment also enables more flexible handling with regard to a distribution of services, in particular a dynamic distribution of services, that is to say a distribution of services that can change over time, and in particular in which such changes can be implemented with little effort. For example, a “service assignment”, ie an assignment of which node can or should perform which service, can be stored in just one node (a third node). If the service assignment changes or should change, the service assignment only needs to be updated in the third node, but not in any other node that might want to access the corresponding service. In this way, a “relocation” of services from one node to another node can be implemented advantageously, whereby the update effort can be kept low.

This implementation also enables a certain degree of anonymity, ie the first node that wants to access a certain service does not have to learn (or does not learn) within the scope of this implementation which node can / should or is executing the requested service. According to one embodiment, the method also includes rejecting a call, in particular at or through a node lying on the routing path between the first node and the second node, in particular in the event that the call or the communication connection is not authorized.

According to this embodiment, a call can be rejected, for example because the “destination node” (the second node) is not available for a requested service, for example because it is no longer performing the requested type of service or because there is no free capacity to perform this service Has.

A call can also be rejected if there is no authorization for this call or the communication connection. This could be used, for example, to implement company-internal access authorizations or to make selected services available to nodes only after a corresponding fee has been paid for this. A corresponding check can be carried out by a node on the routing path, for example by the second node itself, or even by the first node. However, the checking or rejection advantageously takes place at a node located between the first and the second node, for example at a node within a company intranet. This could be centrally responsible for checking and rejecting, in particular with regard to several second nodes.

According to one embodiment, at least one of the first and second nodes is an autonomous transport device, a stationary or mobile manipulator, a stationary or mobile manipulator arrangement, in particular a stationary or mobile robot or a stationary or mobile robot arrangement, or a controller for this, or a combination of two or more of these.

Autonomous transport devices in the context of automatable industrial devices or automatable industrial plants include in particular transport vehicles that can transport objects autonomously, for example within a building, such as a warehouse, or between several buildings / warehouses or on a factory site. In a simple case, these could have a loading platform or a container / container to move the items from one place to another To be transported in place without loading and / or unloading yourself. Such transport vehicles can also be equipped with suitable means for loading and / or unloading, for example a lifting device such as a forklift truck, a gripper arm, etc.

The manipulators, manipulator arrangements, robots or robot arrangements that come into consideration include, in particular, devices that process a workpiece or change its position or position.

A second aspect of the invention relates to a control device for a node, in particular for a manipulator, in particular for a robot, the control device being set up to carry out one of the methods described above.

A third aspect of the invention relates to a computer program that executes one of the methods described above when it runs in a control device as described above.

A fourth aspect of the invention relates to a computer program product with program code which is stored on a machine-readable carrier and which includes a computer program as described above.

In one embodiment, a computer program product can have, in particular, a non-volatile storage medium for storing a program or with a program stored thereon, execution of this program causing a system or a controller, in particular a computer, to use a program here to carry out the described method or one or more of its steps.

Further advantages and features emerge from the subclaims and the exemplary embodiments. This shows, partly schematically:

1: a system or a network structure according to an embodiment of the present invention; Fig. 2: a system or a network structure according to another

Embodiment of the present invention; and

3: a flow chart with method steps of an inventive

Procedure.

Fig. 1 shows a system or a network structure according to an embodiment of the present invention. In Fig. 1, several networks are each shown by dashed rectangles. Each of these networks, viewed individually, forms a homogeneous address space within which individual nodes, which are shown in Fig.

1 are represented by rectangles with solid lines, are clearly addressable, i.e. H. have a unique network address. Within each network there must therefore be an agreement between the participants or nodes about the address assignment.

The nodes shown in FIG. 1 can in principle communicate with one another or exchange data. This communication can take place, for example, by wire (e.g. Ethernet) or wirelessly (e.g. cellular radio, WLAN). However, according to FIG. 1, not every node shown can communicate directly with every other node shown. In the exemplary embodiment in FIG. 1, all possible direct connections are shown by solid lines.

As already mentioned, the present invention is used in connection with the communication between two nodes, at least one of these nodes being an automatable industrial device, installation or control for this. Examples of these have already been mentioned. In the interest of a clear description, it is now assumed in connection with the exemplary embodiment in FIG. 1 that one of the nodes is a robot, without this being intended to apply as a restriction. As a concrete example, the network K-1 shown in FIG. 1 can be the intranet of a robot manufacturer. The network CL-C represents a cloud network, the network CL-S another cloud network, the network C-D a customer network and the network C-T another customer network.

The node K in the network Kl represents an interface with access to the cloud network CL-C. The same applies to the nodes D and T. The two cloud networks Networks are connected to one another via the respective interfaces S and C, whereby the interface C can also communicate with the interfaces K, D and T. The other nodes in the networks Kl, CD and CT cannot communicate directly with the interface C, but only via the respective interfaces K, D and T in their respective networks.

We now assume in the sense of an example that the node 1 within the network C-T is a robot. We also assume that the node 1 in the network K-1 can provide a certain service, for example the diagnosis of a problem that can occur with a (customer) robot, for example with the robot 1 in the customer network C-T. Furthermore, we assume that the node 1 in the network K-1 is actually supposed to carry out a diagnosis of the robot 1 in the customer network C-T, for example in the context of a check that recurs at fixed time intervals or on the basis of an explicit request to do so.

In order to carry out a diagnosis, the node 1 in the network K-1 must now, for example, query sensor data from the robot 1 in the network C-T, i. H. these must be requested and then transmitted. If the addressing in the overall network 50 shown in FIG. 1 were to take place according to conventional addressing methods, then all the nodes shown in FIG. 1 would have to be uniquely globally addressable. In the example shown, this is problematic because there is a node with the identifier “1” not only in network C-T, but also, for example, in networks C-D and CL-C. According to conventional addressing methods, these nodes would first have to be assigned unique addresses.

In contrast to this, according to the invention, such a unique global addressing is not necessary. According to the exemplary embodiment presented according to the invention, several sub-networks may have and maintain a node with the same identifier, here for example “1”. So that node 1 in network Kl can now query sensor data from robot 1 in network CT, node 1 in network Kl - this node is referred to here as the "first node" - sends a call to set up a communication link between the first node (i.e. itself itself) and the robot 1 in the network CT - here as “second Node ”. This call contains routing path information which specifies the routing path from the first node to the second node. In this case, the routing path leads via the node K to the node C, from there to the node T and then to the second node with the identifier “1. The routing path information could, for example, have identifiers of these nodes to be traversed. However, the specification of the identifier K is not necessary because, due to the other specified identifiers, such as C, it is already clear that the routing path leads out of the network Kl and because in the example shown in FIG. 1 the only connection that leads out of the network Kl, via the node K leads. In the present example, specifying the identifiers C, T and 1 would be sufficient to clearly specify the routing path from the first node to the second node. In Fig. 1 this is shown by the dash-dotted line labeled “C / T / 1”.

Instead of an indicator of the nodes to be traversed, the routing information could also have identifiers of the networks to be traversed, in the present example the networks CL-C and C-T, or a mixture of identifiers of the networks and nodes to be traversed. In this sense, in the last example mentioned, the specification of the identifier “1” would also be necessary for the second node so that the routing path to the second node is clearly specified.

In a variant, the specification of the identifier of this second node would not be necessary if, for example, all calls to the robot with the identifier 1 are routed as standard in the network C-T.

The communication link between the first node and the second node can then be set up based on the routing path information. As a payload, for example, a request can be sent from the first node along the routing path to the second node so that the second node transmits sensor data to the first node.

Fig. 1 indicates further examples of communication links by dash-dotted lines. These are briefly explained below. On the far left in FIG. 1, four individual nodes 1, 2, 3 and 4 are shown, which form independent networks, ie their networks contain only a single node. These are connected to the CL-S cloud network, ie they can (only) communicate with it directly.

The routing path from node 4 to node S leads directly from node 4 to node S without this routing path traversing other nodes.

Accordingly, it would be sufficient if the routing path information for specifying this routing path only contained the identifier “S”, as shown in FIG. 1.

A further communication link is indicated at the top left in FIG. 1, which leads from a node with the identifier “1” via the nodes S and C to a further node with the identifier “1” in the network CL-C. For (unambiguous) specification of the corresponding routing path, specification of the identifiers S, C and 1 would therefore be sufficient, as is indicated in FIG. 1 by the string "S / C / 1". In this example, the specification of the identifier S is provided, although the only direct connection to the node S leads from the node 1 at the top left in FIG. The reason for this is the fact that this connection crosses network boundaries.

In contrast, in a further example of a routing path shown in FIG. 1, the specification of the identifiers C and 1 is sufficient to move from the node with the identifier "4" in the customer network CT via the node T to the node C and on to the node 1 in the cloud. Network CL-C. It is not necessary to specify the identifier T because the only connection from the customer network C-T is via the node T.

Further examples of routing paths shown in FIG. 1 relate to the customer network CD. Within this network CD there is another customer network CF. This contains the nodes F, 1, 2, 3 and 4. The only connection from the customer network CF leads via the node F and from there via a node not identified by an identifier to node D. The network CF could, for example, be a network within the customer network CD, all employees of the company to which these networks are assigned have access to the area of the network CD outside the network CF, but only some employees also have access to the network Have CF. A routing path from node 4 to node 3 in network CF is sufficiently defined in that only the identifier “3” is specified in the corresponding routing path information, as indicated at the bottom right in FIG.

The routing path from node 1 within network CF to node C in cloud network CL-C necessarily leads out of network CF via node F, from there via node D to leave network CD, and from there to the node C in the cloud network CL-C. This path is sufficiently defined by specifying only the identifier “C”, because it is clear from the outset that all paths to leave the network C-F and C-D must lead via the nodes F and D. It is therefore not necessary to specify the identifiers F and D in the corresponding routing path information.

Finally, a last routing path example is indicated in FIG. 1, which in a certain way can be viewed as a reversal of the immediately preceding example. Here the routing path should lead from the node C in the cloud network CL-C via the node D in the customer network C-D and the node F in the customer network C-F to the node 2 in the network C-F. In contrast to the previous example, it is now necessary to specify the identifiers D, F and 2 in order to clearly define the routing path, for example using the string “D / F / 2”, as indicated in FIG. 1. Specifying only the identifier “2” would in this case not be sufficient for a clear identification of the routing path, because without specifying the identifier D, starting from node C, it is not clear from the outset that the routing path should lead into network C-D.

2 shows a further example of a network system or overall network 50. The overall network 50 has four (independent) sub-networks. In the interest of a simplified representation, these are not shown separately. A first network has nodes 10 to 13, a second network has nodes 20 and 21, a third network has nodes 30 to 32 and a fourth network has nodes 40 and 41.

Each of these four networks represents a homogeneous address space so that each node within each of these four networks can be uniquely addressed.

Although all nodes shown in FIG. 2 are also different when viewed overall Carrying reference symbols between 10 and 41 does not mean that the overall network 50 also forms a homogeneous address space. The reference signs between

10 and 41 are therefore not necessarily to be regarded as identifiers of the individual nodes in FIG. 2, but have only been assigned in the interest of a clearer description so that two nodes do not have the same reference number. The situation is different with the identifiers of the nodes, which are not shown in FIG. 2, but which can be assigned to the individual nodes in a manner analogous to FIG. 1. Thus, in FIG. 2, too, two different nodes in two different networks can carry the same identifier. For example, nodes 11 and 31 could have the identifier “1”, nodes 12 and 32 the identifier “2” etc. However, in the following description - again in the interest of better clarity - the reference symbols of the nodes are treated as if they were at the same time the identifiers of the nodes.

Analogously to FIG. 1, the direct connections (that is to say those that do not have to lead via further nodes) are shown by solid lines or double arrows. Direct communication is therefore possible between node 11 and node 10. On the other hand, a communication link between node 11 and node 20 must necessarily lead via node 10.

The establishment of a communication connection will now be explained in more detail using an example according to the invention. We now assume that from knots

11 a call is sent in order to set up a communication link between this node 11, which is thus to be regarded as the “first node”, and the node 31, which is to be regarded as the “second node”. This communication link leads from node 11 via nodes 10, 20 and 30 to node 31. The corresponding routing path P is illustrated by an arrow with a dashed line.

In a first step, the first node 11 sends a call to set up the communication connection with the second node 31, for example to request sensor data from node 31. As described with reference to FIG. 1, this call contains the necessary routing path information, that is to say the identifiers of at least the nodes 20 and 30, and possibly also the identifiers of the node 10 and / or the node 31. In the following, we assume that the routing path information includes the (node) identifiers of all on the routing path P having lying nodes, so the identifiers of nodes 10, 20, 30 and 31, for example in the form of a string "10/20/30/31".

Based on the routing information contained in the call, the node 10 forwards the call to the node 20, the latter to the node 30 and the latter finally to the second node 31. The call was thus routed to the second node 31, i. H. a communication link is set up between the first node 11 and the second node 31.

While a particularly simple example of setting up a communication link between a first and a second node was described above, many design options and variants are possible for this. These are described below.

version 1

After the call has been routed to node 31 as described above, it can use the routing information to recognize (for example because its node identifier is the last identifier contained in the routing information) that the call is intended for it as the destination node. The second node 31 can then execute instructions which can be contained in the call in the form of a payload. For example, if the call contains the instruction to collect sensor data and send it back to the first node 11, the second node 31 can collect this data and then a corresponding message, which in turn contains the routing information (albeit in the reverse order), on the return path of the routing path P. send back. This message can be passed on through nodes 30, 20 and 10 on the way there, analogously to the call. Depending on the design, it is not necessary to specify the node identifier of the originally first node 11 in the routing information, for example because node 11 is the only (further) node in the network in which node 10 is located.

According to the exemplary embodiment described so far with reference to FIG. 2 and its further development, each of the messages sent would have to contain the (complete) routing information for the routing path P. Variant 2

In a further embodiment according to the invention, the latter can be avoided in order to save memory space in the routing nodes or in memories connected to them and / or to make the routing of the messages more efficient. For this purpose, the first node 11 can first send a start message. In addition to the routing information, that is to say the identifiers of the nodes on the routing path P, this contains a connection identifier. This identifies the connection to be set up, for example in order to distinguish it from other communication connections set up or to be set up.

The start message is processed after it arrives at node 10. For this purpose, the routing path information is stored in a corresponding memory under the connection identifier, for example in a type of routing table. The start message is then forwarded accordingly to node 20, processed there accordingly and forwarded to node 30. This also processes the start message and finally forwards it to the node 31, where the start message is also processed accordingly. This means that the routing path information is stored in all affected nodes so that the routing path information can be called up again using the connection identifier. This has the particular advantage that further messages belonging to this communication connection do not have to contain the routing path information again. It is therefore sufficient that, for example, the message from node 31 to node 11 only has the connection identifier on the way back. The nodes on the routing path P can then forward this return message accordingly on the basis of the stored routing path information. The same applies to further messages that are sent from the first routing node 11 to the second routing node 31 or vice versa.

Variation 3

In a further advantageous further development, not all of the routing path information is stored for each participating node, but only a relevant part. To illustrate this, let us assume that the routing information contained in the start message has the format “10/20/30/31”, for example. The start message also contains a connection identifier, as described. The start message is now sent from the first node 11 to the node 10. Using the routing information, the latter recognizes that the start message is to be forwarded to node 20. The node 10 therefore forwards the start message to the node 20. In addition, the node 10 stores the connection identifier and, connected to it, the node identifiers of the nodes 11 and 20 in its or an assigned memory. The node 10 can take the node identifier of the node 20 from the routing information. The node 10 can infer the identifier of the node 11 from the fact that it has received the start message from the node 11. Other parts of the routing information, in particular the identifiers of the nodes 30 and 31, are not stored in the routing table in the node 10 because the latter cannot communicate directly with these nodes anyway.

The node 20 proceeds accordingly after receiving the start message and stores in its routing table under the connection identifier the identifier of the node 30 as the identifier of the neighboring node in the destination direction (i.e. towards the destination node 31) and also the identifier of the node 10 as the identifier of the neighboring node in the start direction (i.e. towards starting node 11). In addition, node 20 forwards the start message to node 30, which proceeds accordingly and finally forwards the start message to node 31. This is itself the destination node, so that only the identifier of node 30 has to be stored there under the connection identifier as the identifier of the neighboring node in the start direction.

Further communication connections that have now been set up then no longer need to contain routing path information, but only need to specify the connection identifier. The nodes involved can then use the routing tables stored in them to determine to which neighboring nodes the messages are to be forwarded.

Variation 4

In a further embodiment, the routing path information contained in the start message can be used when it is forwarded to the next neighboring node be shortened. If, for example, the routing path, as described above, leads from the first node 11 via the nodes 10, 20, 30 to the second node 31, the routing path information could in principle be through the string

“11/10/20/30/31” can be displayed. On the other hand, the node identifiers of nodes 11 and 10 do not have to be contained in the start message that is sent from node 11 to node 10. The node 10 knows its node identifier anyway, and the node 10 also recognizes that the start message was sent by the node 11. Because the only connection from the network in which nodes 10 and 11 are included leads via node 10, the start message in this case can only contain the string “20/30/31” as routing path information. When forwarding the start message to node 20, node 10 can shorten the routing path information by the node identifier of node 20 so that the start message to be sent to node 20 then only has the string “30/31” as routing information. Node 20 in turn recognizes that it has received the start message from node 10, so that it can store this information in its routing table. Node 20 proceeds accordingly and shortens the routing path information to “31” and forwards the start message with this shortened routing path information to node 30. This also recognizes that it has received the start message from node 20 so that it can note this information in its routing table. The start message to be forwarded from the node 30 to the node 31 can then contain an “empty path” because the node 31 knows that it has received the start message from the node 30 and has not explicitly pointed this out based on the routing path information contained in the start message must become.

Connection identifier

In particular because the connection identifier mentioned above is used in accordance with some of the embodiments described above for the correct routing of the messages, it is of certain importance. For this reason, it makes sense to ensure that the same connection identifier is not used for two different communication connections. In principle, this could be achieved through a global agreement in all participating networks, but it would be difficult to implement and / or inflexible under certain circumstances. Although possible, assigning a connection identifier would also be an example based on the start node and a time specification may not be unique, because two different nodes in two different networks may have the same node identifier and may send a call at the same time.

According to one embodiment according to the invention, an alternative method for assigning connection identifiers is proposed: The connection identifiers (or at least a part of each connection identifier) are generated randomly, for example by the first node itself. Alternatively, the connection identifier could be generated by another node in response to a corresponding request and assigned to the first node or the desired communication connection. The communication connection can then be set up in accordance with one of the methods described above.

With connection identifiers generated at random, however, it can also happen that two desired communication connections, for example starting from two different nodes, are assigned the same connection identifier. In the exemplary embodiment in FIG. 2, for example, the node 11 could wish to have a communication connection with the node 31 and generate a connection identifier for this purpose, for example “17”. The communication connection is established according to one of the methods described above. At about the same time, or at least while the communication connection between node 11 and node 31 is or is being set up, node 41, for example, wants a communication connection with node 32 and would like to use the connection identifier “17” for this. The node 41 then sends its start message to the node 40, which forwards it to the node 20, as described above. Node 20 establishes that a communication connection with connection identifier “17” already exists in its routing table. In order to avoid a “collision” or to ensure that messages for both communication connections are correctly routed, the node 20 rejects the second start message with the connection identifier “17”. The communication link desired by the node 41 is therefore initially no longer set up. Instead, the node 20 sends a message to the node 40 and this accordingly sends a message back to the node 41 in order to notify these nodes that a communication connection with the connection identifier “17” already exists. Based on this message, the Node 41 in turn randomly assigns a new connection identifier for the desired connection. A new attempt is then made to set up the desired communication connection based on the new connection identifier. Depending on the implementation, it can be relatively unlikely that this retry will be rejected again. If this should be the case, however, this part of the method is repeated until the communication link between node 41 and node 32 can be set up without collision.

Terminating a communication connection

A further development that can be used in connection with all of the methods described above relates to the termination or disconnection of an established communication link. This is again described with reference to FIG. 2, again with reference to the example according to which node 11 had requested sensor data from node 31. The clearing down of the communication link between the first node 1 and the second node 31 can be initiated, for example, by a finish message. This could for example be sent by the second node after it has determined that it has transmitted all the requested sensor data, or by the first node 11 after it has determined that no further sensor data is required from the node 31.

We now assume that the second node 31 sends the finish message and that the dismantling begins with the node 31. The finish message is sent along the routing path P like other messages belonging to the corresponding communication link, namely in this example from node 31 via nodes 30, 20 and 10 to node 11. Each node involved marks each node involved by sending the finish message in its routing table the communication connection or the connection identifier as deactivated or deletes the corresponding entry, in particular together with the stored node identifiers of the preceding node and / or successor node. The latter in particular so as not to unnecessarily block storage space. In any case, the corresponding connection identifier is then available for future desired communication connections. Thus the probability becomes a Collision of communication links with the same link identifier minimized.

If, as an alternative, the finish message is sent by the first node 11, the clearing takes place in the reverse order. It is also possible for the finish message or the disconnection to originate from another node, for example node 20. The disconnection of the communication link then propagates in both directions, starting from node 20, that is to say in the direction of first node 11 and the second node 31.

Initial call

In a further refinement, an additional step takes place before a call is sent to set up a communication connection. This is again described with reference to FIG. As a specific application example, it is now assumed that node 11 represents a robot controller. The robot controller 11 would like to have a path planning worked out for a new production process, for example. While this can be done locally / internally with many robots, it may make sense to have a path planning carried out by another node, especially in the event that the path planning is very computationally intensive and the robot controller 11 does not itself have sufficient computing capacity. Node 11 could now request the path planning from a specific node in the system shown in FIG. 2, by means of a call as described above, ie by specifying the corresponding routing path information. According to the present variant, however, the node 11 does not know which node can / should carry out the path planning. Accordingly, at least initially, he cannot send a call with the required routing path information. Instead, the system can be set up in such a way that, for example, it is stored in node 10 which node can / should carry out path planning in response to a corresponding request. In order to request path planning, the node 11 now sends a corresponding message, which in the present case is called an “initial call”. This initial call is sent to node 10 and is recognized by it as a path planning request based on its format or content. The initial call can, for example, have the content “rail planning service”. A table can now be stored in node 10 which indicates that node 13 is responsible for path planning in the same network. Accordingly, a call is generated and sent which contains the routing path information so that communication between node 11 and node 13 can be set up in the same network. It should be noted here that the call is not necessarily sent from node 11. Because it is generated on the basis of the table stored in node 10, the call could be sent by node 10, for example. As a further development, the node 11 does not even have to know or become aware that the node 13 is the target node (or the second node) in the communication connection, because the node 11 communicates directly only with the node 10 and it communicates with the node 11 It is not, or not necessarily, of relevance through which node the path planning is actually carried out.

The variant described last also enables, for example, a transparent relocation of services. For example, due to a change in responsibilities, the path planning service, which was carried out by node 13 up to a certain point in time, could have been “moved” to another node. This could for example be another node in the same network, for example node 12, or a node in another network, for example node 31, which for example belongs to a network of another company. In order to implement this move, it is sufficient that the table in node 10 is updated so that node 13 is no longer noted under “path planning service”, but rather node 12 or node 31. This “move” becomes the Node 11, which requests the path planning service, is not known or not necessarily known. As described above, node 11 sends an initial call to node 10, whereupon the latter sends the actual call to set up the communication link, which is now not set up with node 13 but, in the example mentioned, with node 12 or the node 31. This is particularly advantageous if several nodes (for example several robots in a company) want to access the same service. Instead of storing new path information for the relocated service in each individual node / robot etc., an update for only one node (node 10 in the example mentioned) is sufficient to correctly route future requests for the corresponding service. Access restrictions

Another embodiment is again described with reference to FIG.

According to this embodiment, access options can be restricted. This is of particular interest if the networks concerned do not all belong to the same company, for example. For example, an access control can be installed in node 30 in FIG. 2, which controls access to the network to which nodes 30, 31 and 32 belong. In this example, every call which is routed from node 20 to node 30 is analyzed by the latter. If the appropriate authorization is available, the call can be forwarded. Otherwise it will be rejected.

Such an access control does not necessarily have to be installed in a node belonging to the network to be controlled. Access control for the network to which nodes 30, 31 and 32 belong could therefore also be installed in node 20, for example.

The access control presented here can protect networks against unauthorized access. Such an access control can also ensure that nodes, for example, only have access to services for which a corresponding fee has been paid.

3 shows a flow chart with method steps of an exemplary embodiment of a method according to the invention. After the method has started 100, a call to set up a communication connection is sent in a step 101. The call has routing path information which specifies the routing path from a first node to a second node. In a further step 102, the communication connection between the first node and the second node is set up based on the routing path information. The method can then end (step 103).

Further details of embodiments of a method according to the invention have already been described above. Embodiments of the present invention have various uses. The following applications may be mentioned as non-limiting examples.

example 1

A user has a user interface with which he can manage a specific robot user application. With the present invention, this user interface can be used essentially anywhere: directly connected to the robot; on a WiFi network in the factory where the robot is located; on an intranet of the company to which the robot belongs; or over the public internet.

Depending on the design and other factors, the routing path or the routing path information may or may not already be known to the user or to the device (computer, laptop, tablet, smartphone, etc.) that the user is using as an interface. Such factors can in particular be the current “location” of the device used as an interface and / or the (direct) connection to a specific network.

If the routing path or the routing path information is not already known, these can be determined. There are several possibilities for this, which can also be used in other examples or embodiments:

(a) Path in the file system: There is a root in the network that can be directly addressed from anywhere (e.g. a node in the cloud network of the robot manufacturer) and there is a logical topology, e.g. cloud / customer / factory / robot

(b) Interactive browsing: The user searches the network himself starting from his local node. The interaction is then similar to that in a filesystem browser. The user navigates step by step from one node to the next (neighboring) node and from there to the next (neighboring) node, etc. When the user has finally found his destination, the path found in this way corresponds to the routing path sought. (c) Telephone book / search service: There are directly accessible services that save the routing paths and make them searchable on the basis of additional information. Services / robots can log on to such a service with any information and can be found. For example, the customer first gives a service (for example on cloud / kundel / search) a robot name as a search term and receives the routing path as a response from this service.

(d) Domain name system (this can correspond to what is described in the “Initial Call” section or build on it): The target service / robot has an abstract name. This abstract name is known to the user. In addition, the abstract name and the current routing path leading to the target service / robot are stored in directly accessible services. If a service / robot "moves", the stored information is updated. The user can then access the target service / robot using the abstract name known to him.

Example 2

A customer's robot is equipped with additional sensors for a customer application. Customer-specific services are developed that read out the sensor data and run on the robot. These customer-specific expansion services are connected on the robot to the overall network 50 (shown for example in FIG. 1) and are therefore also available, for example, in a cloud network without changes to the communication from the customer to the network of the robot manufacturer being necessary.

Example 3

Reading out sensor data has already been mentioned. For example, an employee of a robot manufacturing company can use the company's intranet to query sensor data relating to a robot that is in another network as part of customer service. Sensor data can also be automatically reported from the customer network to the network of the robot manufacturer company. This data can be checked for anomalies, for example, in the customer network or in the network of the robot manufacturer. In the event of a significant deviation from specified values, additional Sensor data are requested, for example to enable a diagnosis of a problem with respect to the robot.

Example 4

It has already been described that, for example, path planning services or other services that could be of interest in connection with an automatable industrial plant can be outsourced, i. H. are not provided directly by or on the industrial system. As a result, storage capacity and / or computing capacity can be saved in the system or in several systems, which can promote more cost-effective and / or resource-saving production. The corresponding services can then be provided centrally, i.e. for several systems, by a node in another network, for example. In addition, the provision of these services can be dynamically shifted to other nodes. In particular, if this relocation is carried out "transparently", it may only be necessary to update a route table or the like at a single node, and not at all nodes that use the service in question, as already described.

Although exemplary embodiments have been explained in the preceding description, it should be pointed out that a large number of modifications are possible. It should also be pointed out that the exemplary designs are only examples that are not intended to restrict the scope of protection, the applications and the structure in any way. Rather, the preceding description provides a person skilled in the art with guidelines for the implementation of at least one exemplary embodiment, whereby various changes, in particular with regard to the function and arrangement of the described components, can be made without departing from the scope of protection as it emerges from the claims and these equivalent combinations of features. Reference list

1-4 nodes / identifiers

C, D, F, K, S, T nodes / identifiers K-l network (intranet robot manufacturer)

CL-C, CL-S (cloud) networks C-D, C-F, C-T (customer) networks 10-13 nodes 20-21 nodes 30-32 nodes

40-41 nodes 50 total network

100-103 process steps

Claims

1. A method for setting up a communication link between a first node (11) in a first of at least two networks and a second node (31) in a second of the at least two networks, at least one of the first and second nodes (11, 31) is an automatable industrial device or an automatable industrial plant or a control therefor, the at least two networks individually each forming a homogeneous address space, the at least two networks together not forming a homogeneous address space, the method having the following steps:

Sending (101) a call to set up the communication link between the first node and the second node, the call having routing path information (10/20/30/31) which describes the routing path from the first node (11) to the second node (31), the routing path information (10/20/30/31) at least one identifier (10, 20, 30, 31) of each of the networks or nodes to be traversed on the routing path, but not necessarily an identifier of the first Network, have; and

Establishing (102) the communication link between the first node (11) and the second node (31) based on the routing path information (10/20/30/31).

2. The method according to claim 1, wherein the routing path information (10/20/30/31) has an identifier of the second node (31).

3. The method according to claim 1 or 2, wherein the call further comprises a connection identifier.

4. The method of claim 3, wherein the communication link between the first node (11) and the second node (31) has at least two messages and wherein all messages of this communication link between the first node (11) and the second node (31) this link - Have an identifier.

5. The method according to claim 4, wherein at least part of the connection identifier is generated or allocated randomly, in particular by the first node (11), and in particular where in the event that the connection identifiers of a first and a second call match and in particular in the event that the routing paths of the first and second call lead via the same node (20), the second call is rejected and / or a new connection identifier is generated for the second call or allocated to it.

6. The method according to any one of claims 3 to 5, wherein for one or the first call in each on the routing path between the first node (11) and the second node (31) lying node or an associated storage device the associated connection identifier and a Node identifier of a previous node and a node identifier of a successor node are stored, the previous node being a node on the routing path, in particular a node immediately adjacent on the routing path, in the direction of the first node (11), and the successor node being a Node lying on the routing path, in particular a node lying directly adjacent on the routing path, in the direction of the second node (31).

7. The method according to any one of claims 3 to 6, wherein the routing path information (10/20/30/31) and the associated connection identifier of a call are contained in a start message.

8. The method according to claim 7, in particular as a function of claim 6, wherein messages of this call following the start message have the associated connection identifier, but not the routing path information (10/20/30/31), the messages relating to the start message The following messages are preferably routed using the node identifiers stored on the routing path between the first node (11) and the second node (31) or the associated storage device.

9. The method according to claim 6, or according to one of claims 7 to 8 in direct or indirect dependence on claim 6, wherein when the call is terminated in the nodes lying on the routing path between the first node (11) and the second node (31) or the associated storage device, the stored associated connection identifier and / or the stored node identifier of the previous node and / or the stored node identifier of the successor node are deactivated, in particular deleted.

10. The method according to any one of the preceding claims, wherein the communication link is a bidirectional communication link.

11. The method according to any one of the preceding claims, further comprising: before sending the call, sending an initial call by the first node (11), wherein the initial call has no or only incomplete routing path information, wherein based on the initial Calls the routing path information (10/20/30/31) required for the call can be determined, in particular by a node different from the first node (11), in particular with the aid of an association between information contained in the initial call and the Routing path information (10/20/30/31).

12. The method according to any one of the preceding claims, further comprising rejecting a call, in particular at or by one on the routing path between the first node (11) and the second node (31) lying nodes, especially in the event that the call or the communication link is not authorized.

13. The method according to any one of the preceding claims, wherein at least one of the first and the second node (11, 31) is an autonomous transport device, a stationary or mobile manipulator, a stationary or mobile

Manipulator arrangement, in particular a stationary or mobile robot or a stationary or mobile robot arrangement, or a control for this, or a combination of two or more of these.

14. Control device for a node, in particular for a manipulator, in particular for a robot, the control device being set up to carry out a method according to one of the preceding claims.

15. Computer program that executes a method according to one of claims 1 to 13 when it runs in a control device according to claim 14.

16. Computer program product with program code which is stored on a machine-readable carrier and which comprises a computer program according to claim 15.