Methods and systems for providing data from an internal data processing system of an industrial plant to an external data processing system

The invention relates to a method for providing data from an internal data processing system of an industrial plant to an external data processing system, wherein the internal data processing system comprises at least one industrial edge de vice and at least one, particularly a plurality of, industri al machine(s), e.g. machine tools, in particular cutting, boring, grinding, shearing machine tools.

Furthermore, the invention relates to a data processing mod ule for providing data from an internal data processing sys tem of an industrial plant to an external data processing system, wherein the internal data processing system comprises at least one industrial edge device and at least one indus trial machine.

Moreover, the invention relates to an industrial edge device and to an industrial data processing system with at least one above-mentioned data processing module.

In addition, the invention relates to a computer program com prising instructions to cause the above-mentioned industrial data processing system to execute the steps of the above- mentioned method and to a computer-readable medium having stored thereon such computer program.

An industrial network environment can include an ecosystem - hereinafter referred to as - external data processing system and an industrial data processing system - hereinafter re ferred to as - internal data processing system. The internal data processing system can use the external data processing system's infrastructure in order to perform certain tasks. Industrial/internal data processing system comprises or con sists an industrial or internal computer network (networks) of the industrial plant or of a plurality of industrial plants. The external data processing system comprises or is a computer network, e.g. a public computer network, which is external with respect to the industrial or internal computer network.

Various data of an industrial plant, which can be a part of the industrial/internal data processing system is regularly stored and made available in local and/or internal industrial plant networks. These networks are commonly isolated from public networks, because the stored data such as machine, production, planning or scheduling data is sensitive or con fidential and can contain information about production meth ods and processes. This data is routinely used to control, monitor and supervise production flows and processes. There fore, it can be of advantage not to disclose or to transfer such data to external entities in the external data pro cessing system, i.e. entities outside of the private indus trial plant networks, e.g. intranets. Conversely, a direct access to the data by the external entities is often undesir able, because it is associated with certain risks.

For the reasons above there are technical measures that are intended to ensure safety of the internal or confidential da ta in the private industrial plant networks. An example of such technical measure is a firewall, e.g. a network-based firewall, which is positioned, e.g. on a gateway between the local industrial plant network and a public (untrusted) net work, such as Internet or a cloud. A further technical meas ure can involve a type of a tunneling protocol, such as VPN tunneling protocol, in order to extend the internal/private industrial network across the public network and to enable communication between devices across the public network, as if these devices were directly connected to the internal in dustrial network in order to intercommunicate with and, in particular, to exchange data with the internal industrial network.

It is known from the prior art to arrange industrial edge de vices on the "edge" between the private industrial plant net work and the external/public network. The industrial edge de vices are located within the internal industrial network and, on the one hand, provide (at least some) resources for data processing within the internal industrial network and, on the other hand, provide the data from the internal data pro cessing system to an external entity, such as cloud-based so lution, an edge backend device or alike. The industrial edge devices can be also adapted to acquire this data from one or more industrial machines, e.g. machine tools, and/or one or more processes of the industrial plant. Conversely, the in dustrial edge devices are adapted to receive data and com mands from external entities and to apply or implement these in the internal industrial network.

As long as the provider of the external entity or of the ex ternal data processing system or of the entire ecosystem and the operator of the internal industrial network are identical or there exist a special relationship of loyalty and confi dentiality between them, the data exchange between the exter nal and the internal data processing systems can be planned and carried out amicably.

It will be appreciated by the skilled person that the term "internal data processing system" means an industrial data processing system, which is internal with respect to the en tire ecosystem, e.g. to the external data processing system. It will be appreciated that such industrial data processing system can be realized as hard- or software or, more often, as a combination of hard- and software, wherein the hard- and/or software of such "internal data processing system" can e.g. belong to the same legal entity or to a group of associ ated legal entities, e.g. to a corporate group. A problem arises if certain industrial plant related data, e.g. production-related data is not allowed to leave the in ternal, e.g. local industrial network but, at the same time is needed to be provided to the external entity or entities, in order to enable business models, optimizations and alike.

Such situation is possible, if for example an original equip ment manufacturer provides an industrial machine to an opera tor of an industrial plant and the parties agree on a so- called pay-per-use payment model, where the operating time or the number of production operations or alike shall be remu nerated. The original equipment manufacturer and the owner of the industrial machine has, therefore, a legitimate interest in obtaining data, on a regular, e.g. monthly basis, about the use and the utilization of the industrial machine, in or der to correctly bill the operator and to monitor all mainte nance related events, e.g. overloading. The operator of the industrial plant, however, is not willing to expose himself by giving away sensitive data and maybe disclosing confiden tial information such as information about production cycles, order peaks and stagnations, types and design of produced products and alike. Therefore, there is a conflict of inter est between the provider of the external data processing sys tem and the operator of the internal data processing system.

It will be appreciated by the skilled person that the opera tor can be an automated apparatus such as a robot or an arti ficial intelligence, which e.g. can execute and monitor apps on one or more industrial machines.

It will be appreciated by the person skilled in the art, that the provider of the external data processing system and the provider of the entire ecosystem can be different entities. The provider of the external data processing system can be for example a provider of an application software - app - for industrial machines, which software for example, based on the utilization data of the machines optimizes their usage. This app can be stored and run in a cloud, which is a part of an ecosystem, provided by yet another party.

It will also be appreciated, that the provider of the entire ecosystem, the provider/owner of the internal data processing system, e.g. an application software - app - and the operator can be different entities, in particular, different legal en tities. The operator can run an app from the ecosystem on the industrial edge device or devices located in the internal da ta processing system.

Accordingly, there is a need to provide a communication scheme suitable for an untrusted industrial setting.

In order to achieve the objective mentioned above, the pre sent invention provides a method for providing data from an internal data processing system of an industrial plant, par ticularly of an automated industrial plant to an external da ta processing system. The internal data processing system comprises at least one industrial edge device, at least one, particularly a plurality of, industrial machine(s) and at least one data processing module. The at least one data pro cessing module is provided with a, for example certified, in particular, certified by a third trusted party, user-defined data filter. The method comprises: generating a plurality of data packets, e.g. unencrypted data packets, from or based on a data related to the at least one industrial machine by means of the at least one industrial edge device, and signing each data packet of the plurality of for example unencrypted data packets with a first digital signature, in order to pro duce a plurality of signed data packets, particularly a plu rality of the signed read-only data packets by means of the at least one industrial edge device, in particular, by means of each of the industrial edge devices. The method further comprises reading each signed data packet of the plurality of the signed data packets and, while reading, either letting through those signed data packets that pass the user-defined data filter or rejecting those signed data packets that fail to pass the user-defined data filter. This is done by means of the data processing module. The method also comprises sending all those data packets that have passed the user- defined data filter towards the external data processing sys tem.

Signing with the first digital signature can be performed (locally) on a corresponding industrial edge device. In some embodiments the signing with the first digital signature can be performed remotely, e.g. by a software component, e.g. a license manager, which is not co-located with the industrial edge device(s) and/or the industrial plant and can be ac cessed by the industrial edge device(s) in order to perform the signing with the first digital signature.

The data related to the at least one industrial machine can be for example (raw) data extracted/derived from the at least one industrial machine by means of the at least one industri al edge device, which is, in particular, associated with the at least one industrial machine.

Therefore, in some embodiments, generating the plurality of data packets based on a data related to the at least one in dustrial machine can comprise a sub-step of extracting data from the at least one industrial machine.

In an embodiment at least some of the data packets, more par ticularly, all data packets are logfiles.

The data related to the at least one industrial machine can also be a data generated on the at least one industrial edge device. Such data can be based e.g. on edge platform data and/or edge app infrastructure data and/or billing data and/or licensing data etc. Such data can be produced/- generated on the at least one industrial edge device by ana lyzing and processing data from the at least one industrial machine and/or from another (raw) data source within the in ternal data processing system, e.g. data generated by work- flows through software applications (edge device apps) being processed by edge device or edge devices. Such data can be generated by an edge device app that receives data from an other edge app, which receives its data from an industrial machine.

In some embodiments the data related to the at least one in dustrial machine can be based on an infrastructure data, e.g. edge logfiles or on information about how frequently an edge device or an app on the edge had a malfunction or on billing information for pay-per-use of the edge apps etc.

Therefore, in some embodiments, the plurality of data packets can be generated based on a billing data and/or licensing da ta related to the at least one industrial machine. The bill ing data and/or licensing data can be related to the use of the industrial machine itself or to the use of an app, which can some tasks and process some functions related to this in dustrial machine.

The signing of the data packets of the plurality of, in par ticular unencrypted, data packets with the first digital sig nature can be performed by means of the at least one indus trial edge device, on which this data packets are generated. In this case the signing with the first digital signature is done by the ecosystem provider, e.g. manufacturer of the in dustrial edge devices.

It will be appreciated by the person skilled in the art that the above steps can be repeated for any industrial edge de vice.

In an embodiment the letting through those signed data pack ets that pass the user-defined data filter can comprise signing these signed data packets with a second digital sig nature (before sending them towards the external data pro cessing system), in order to produce double-signed data pack ets. This increases the data authenticity and end-to-end data integrity. Signing by the second digital signature can be performed e.g. by the above-mentioned operator of the inter nal data processing system, wherein the operator can perform the signing from the ecosystem, while e.g. running an app or apps on one or more industrial edge devices located in the internal data processing system.

In some embodiments signing with the second digital signature can be performed not on the industrial edge devices but by means of another part of the internal data processing system, for example by a certificate authority. Such certificate au thority does not have to be co-located with the industrial plant. E.g. it can be realized as a software in the cloud- part of the internal industrial network of the internal data processing system.

In some embodiments the method can further comprise validat ing the first digital signature of each data packet or noting the packet's absence by means of the external data processing system.

In some embodiments the method can further comprise storing a first key at the external data processing system, wherein the first key corresponds to the first signature and is used to validate it.

In some embodiments the method can further comprise encrypt ing each signed data packet by means of the at least one in dustrial edge device.

In some embodiments the internal data processing system can further comprise at least one gateway component and the meth od can further comprise sending all those data packets that have passed the user-defined data filter towards the external data processing system through the at least one gateway com ponent. In an embodiment, in which the data processing module signs the signed data packets with the second digital signature, the sending all double-signed data packets towards the exter nal data processing system through the at least one gateway component can comprise following partial steps

- sending all double-signed data packets to the at least one gateway component;

- validating the second digital signature of each of the double-signed data packets by means of a second key, wherein the second key corresponds to the second digital signature;

- sending all validated double-signed data packets to the external data processing system.

In some embodiments the second key can be interrogated by the gateway component, so that the validating the second digital signature can be performed by means of the gateway component. It can be for example stored on the at least one gateway com ponent itself or at some different place within the internal data processing system but not within the industrial plant, which can be accessed by the at least one gateway component.

In some embodiments the second key can be stored within the internal industrial network but not on a physical device within the industrial plant. E.g. the second key can be stored in a cloud (a cloud-part of the internal industrial network, which is a part of internal data processing system), such that validating the second digital signature can be per formed for example by an app in the cloud.

In an embodiment the sending all double-signed data packets towards the external data processing system through the at least one gateway component can further comprise

- encrypting each of the double-signed data packets.

In one of the embodiments the user-defined data filter can be implemented as a whitelist that contains allowed terms de tectable in the content of the data packets, in particular, by finding the terms by searching for the term's text and/or applying image classification and/or object detection and/or mapping the objects detected in images to one or multiple terms.

In one of the embodiments the user-defined data filter is adapted to filter, e.g. to whitelist, the data packets based on the semantic content of the data, which data is contained in the data packets.

In an embodiment the data filter can be implemented as a whitelist that contains matchmaking patterns, in particular, regular expressions, particularly those that can be applied to the content of the data packets.

In order to achieve the objective mentioned above, also a da ta processing module for providing/filtering/whitelisting da ta from an internal data processing system of an industrial plant to an external data processing system is provided, wherein the internal data processing system comprises at least one industrial edge device and at least one (and in, particular, a plurality of) industrial machine(s), wherein the at least one industrial edge device is adapted/configured to

- generate a plurality of (in particular unencrypted) data packets related to the at least one industrial machine, and

- sign each data packet of the plurality of the data pack ets with a first digital signature, in order to produce a plurality of signed data packets, wherein the data processing module is provided with at least one, particularly certified, even more particularly certified by a third trusted party, user-defined data filter and adapted/configured to

- read each signed data packet of the plurality of the signed data packets and, while reading, apply the at least one user-defined data filter to it; - let through those signed data packets that pass the us er-defined data filter or reject those signed data pack ets that fail to pass the user-defined data filter;

- send all those data packets that have passed the user- defined data filter towards the external data processing system.

In an embodiment the data processing module can be also con figured to sign those of the signed data packets that pass the at least one user-defined data filter with a second digi tal signature, in order to produce double-signed data packets and send all double-signed data packets towards the external data processing system.

In order to achieve the objective mentioned above, also an industrial data processing system is provided, which can be connected to an external data processing system, in order to exchange data with the external data processing system and comprises at least one industrial edge device, at least one, particularly a plurality of, industrial machine(s) and the at least one above-mentioned data processing module.

In an embodiment of the invention, the industrial data pro cessing system can be realized as a combination of hard- and software components. In an embodiment the at least one indus trial edge device of the internal data processing system can comprise at least one hardware controller configured to en crypt the data packets. In particular, the industrial data processing system can comprise a hardware controller, for ex ample an FPGA (Field Programmable Gate Array).

The above and other objects and advantages of the invention will be apparent upon consideration of the following detailed description, taken in conjunction with the accompanying draw ings, in which like reference characters refer to like parts throughout, and in which: FIG 1 is a block diagram of an example industrial network environment, in which embodiments with a plurality of industrial machines and industrial edge devices can be implemented;

FIG 2 is a flow diagram of an example of a method for providing data from an internal data processing sys tem of an industrial plant to an external data pro cessing system;

FIG 3 is a block diagram of an example industrial network environment with a single gateway component between an internal and an external network; and

FIG 4 is a block diagram of an example industrial network environment and a third trusted party for providing key pairs.

Turning to FIG 1, a block diagram of an example industrial network environment is shown. The industrial network environ ment can include an ecosystem - hereinafter referred to as - external data processing system EDP and an industrial data processing system - hereinafter referred to as - internal da ta processing system IDP. The internal data processing system IDP can use the ecosystem's infrastructure in order to per form certain tasks. More particularly the external data pro cessing system EDP can be a cloud-based external data pro cessing system EDP (a cloud-based ecosystem). The external data processing system EDP can comprise a network environment - hereafter referred to as external network - for enabling communication for the entities of the ecosystem and computing devices and/or systems, e.g. backend devices EB or systems that are adapted to analyze data and perform analytics.

The internal data processing system IDP can comprise an in dustrial plant PL. It will be appreciated that the internal data processing system IDP can be a part of - a local data processing system of the industrial plant PL or co-located with the industrial plant PL. The industrial plant PL can be an automated industrial plant. The internal data processing system IDP can include a plurality of industrial machines IM1, IM2, ... IMn, such as motors, robots, industrial equip ment, automatization devices, and (a plurality of) industrial edge devices EDI, ED2, ...EDn, wherein each industrial edge device can be connected to a single industrial machine IM1, IM2, ... IMn and can intercommunicate with it in order to ex change data. Different industrial edge devices EDI, ED2, ...

EDn can be connected to different industrial machines IM1,

IM2, ... IMn. The industrial edge devices EDI, ED2, ...EDn and the industrial machines IM1, IM2, ... IMn can be co-located with the industrial plant PL (FIG 1).

The industrial plant PL can be a manufacturing and/or testing facility, an industrial power plant, a facility for industri al plant engineering etc.

Industrial edge devices EDI, ED2, ...EDn can be connected (for exchanging data, e.g. wirelessly) to a controller of the ma chine tool IM1, IM2, ... IMn within the industrial plant PL and, therefore, co-located with the physical systems in the industrial plant environment. Industrial edge devices EDI, ED2, ...EDn can be adapted to extract data from at least one of the plurality of the industrial machines IM1, IM2, ... IMn. Moreover, the industrial edge devices EDI, ED2, ...EDn can be adapted to provide (at least some) resources for data pro cessing within the internal data processing system IDP of the industrial plant PL or - shortly - within the internal net work of the industrial plant PL, for example within a factory LAN.

The industrial edge devices EDI, ED2, ...EDn can be configured to produce/generate data packets DPI, DP2, ...DPn, for example unencrypted data packets. The data packets can be generated (on industrial edge devices EDI, ED2, ...EDn) based on a data related to one or more industrial machine(s) IM1, IM2, ... IMn. In an embodiment at least one or a part of or all of the in- dustrial edge devices EDI, ED2, ...EDn can comprise a hardware component configured to encrypt the data packets DPI, DP2, ... DPn. In particular, such hardware component can be a Field Programmable Gate Arrays (FPGA) controller. Such hardware components can increase the computing power of the industrial edge devices, because the resources needed for the encryption may be made available for performing other tasks, such as running apps.

For example the data packets can be generated based on a raw data, which is extracted from the industrial machines IM1,

IM2, ... IMn, and/or on a data related to the raw data, but was already analyzed and/or processed by one or more industrial edge device(s) EDI, ED2, ...EDn, for example by means of a software application (of an edge app). Such app can for exam ple use hardware resources of one or more of the industrial edge device(s) EDI, ED2, ...EDn to process the raw data in or der to generate the data packets DPI, DP2, ...DPn. The data related to one or more industrial machine(s) IM1, IM2, ... IMn can be also provided to the one or more industrial edge de vice (s) EDI, ED2, ...EDn by the internal data processing sys tem IDP, e.g. by a software component of the internal data processing system IDP, e.g. by another app, e.g. edge app, which can be processed on the same industrial edge device(s). The later type of data be based on the data workflow in the internal data processing system IDP, e.g. between the apps deployed on one or more industrial edge devices EDI, ED2, ... EDn. It will be appreciated that the above-mentioned software components, such as apps or edge apps can be processed re motely, e.g. from the cloud, i.e. the user running a particu lar app does not have to be co-located with the industrial plant PL. In this way the internal data processing system IDP can use the infrastructure of the ecosystem.

The data related to the at least one industrial machine IM1, IM2, ... IMn can be based e.g. on data or Edge platform and/or Edge app infrastructure and/or billing and/or licensing data etc. FIG 1 illustrates an example, where each single industrial edge device EDI, ED2, ...EDn can be connected to/associated with a single industrial machine IM1, IM2, ... IMn correspond ing to that industrial edge device to intercommunicate and, in particular, exchange and/or receive/extract data from it. For sake of simplicity FIG 1 depicts a situation, in which a single data packet DPI, DP2, ...DPn is generated on each in dustrial edge device EDI, ED2, ...EDn. It will be, however, appreciated that a plurality of the data packets can be gen erated on each of the industrial edge devices EDI, ED2, ...EDn with time.

The industrial edge devices EDI, ED2, ...EDn can be adapted to provide the data packets DPI, DP2, ...DPn directly, i.e. with out any further processing and/or passing through further de vices, to an external entity in the external data processing system EDP (in the ecosystem). Such entity can be situated in the external data processing system EDP, e.g. in the cloud, and can be for example an external software application - an app - or a cloud-based solution or an edge backend EB. In this situation each industrial edge device EDI, ED2, ...EDn serves as a gateway agent, for example as an IoT gateway.

This can be the case, when the entire industrial network en vironment, including the communication channel between the internal and the external data processing system, e.g. edge- to-cloud communication channel, can be completely trusted.

The industrial edge devices EDI, ED2, ...EDn can be also adapted to encode/encrypt the data packets DPI, DP2, ...DPn before sending them via one or more uplinks UL1, UL2, ...ULn to the external data processing system EDP, e.g. to the ex ternal entity, e.g. an edge backend EB, in the external data processing system EDP. This can be of advantage, if the com munication channel between the internal data processing sys tem IDP and the external entity in the external data pro cessing system EDP or the external data processing system EDP itself cannot be fully trusted. The data packets DPI, DP2, ...DPn can have different struc ture. In general, they can be unstructured, semi-structured and structured. For example, the data packets DPI, DP2, ...DPn can be logfiles.

In some scenarios, especially in untrusted situations, the data packets DPI, DP2, ...DPn from the industrial edge devices EDI, ED2, ...EDn can go through a separate gateway device GW (FIG 3 and FIG 4) before they leave the internal network of the internal data processing system IDP, e.g. the local net work of the industrial plant PL. This can be motivated by se curity reasons, because it is more feasible to monitor one upload link instead of n, where n is the number of the edge devices.

Before being provided to the external entity, such as backend device EB in the external data processing system EDP the data can be as well pre-processed/ processed, e.g. encrypted and/or signed by the industrial edge devices EDI, ED2, ...EDn. Moreover, the industrial edge devices EDI, ED2, ...EDn can be adapted to receive data from the external data processing system EDP. Such data can comprise data from an external ap plication software - apps, and/or commands from an external entity and/or firmware updates etc. The industrial edge de vices EDI, ED2, ...EDn can be adapted to implement/apply the received data accordingly within the internal data processing system IDP of an industrial plant PL.

As described above there are, however, situations, where only a very low level of trust is allowable, because sensitive and/or confidential data is contained in the data packets DPI, DP2, ...DPn.

For that reason, at least one data processing module DPM is provided. The data processing module DPM is contained in the internal data processing system IDP and can be a separate network component or a software module, particularly a soft ware component, more particularly a plugin, which can be in- stalled on a network component, such as a proxy-server, a gateway component or an industrial edge device of the inter ^¬ nal data processing system IDP and can be implemented by this network component. More particularly, the data processing module DPM can be co-located with the industrial plant PL.

The at least one data processing module DPM is provided with a user-defined data filter DF. The user-defined data filter DF can be certified, for example certified by a third trusted party. The term "user-defined" means that the data filter DF is defined by the user of the provided product, in the cur ^¬ rent context - by the operator of the industrial plant PL.

The user-defined data filter DF contains information about, which data is allowed to leave the internal data processing system IDP and which is not. The user-defined data filter DF can be a whitelist. Therefore, in an embodiment the data pro ^¬ cessing module DPM can be a whitelisting plugin with a user- defined data filter DF in form of a whitelist. The user- defined data filter DF and/or data processing module DPM can be certified. The certification can be performed by the pro ^¬ vider of the industrial machine IM1, IM2, ... IMn and of the industrial edge device EDI, ED2, ...EDn and/or by the provider of the external entity, e.g. by the provider of the applica ^¬ tion software or by an independent trusted third party TC (cf. FIG 4), such as German technical inspection association (Tt)V).

For example, the user-defined data filter DF can comprise a list of words and/or matchmaking patterns (e.g. regular ex ^¬ pressions, also called "searchpatterns", matching allowed terms) and/or object descriptions (if data packets contain images). The data processing module DPM can for example use neuronal networks for object recognition in recorded images and, when an object, e.g. a spindle or a spindle nose, is recognized, compare it with the content of the user-defined data filter DF. Then, if for example the form of a spindle is an industrial secret, then the corresponding data packet con- taining a picture of the spindle can be blocked by the data processing module DPM.

Turning to FIG 2, an example of a method for providing data from the internal data processing system IDP of the industri al plant PL to the external data processing system EDP is il lustrated.

In the first step SI of the method raw data can be extracted from the at least one industrial machine IM1, IM2, ... IMn.

From this data a plurality of data packets DPI, DP2, ...DPn can be produced - step S2. At this stage the data packets can be encrypted or unencrypted. Each data packet DPI, DP2, ...DPn can be signed with a first digital signature PR1, in order to produce a plurality of signed data packets - step S3. The steps SI to S3 can be performed by means of the industrial edge devices EDI, ED2, ...EDn. Step S3 assures that the data comes from the industrial machines IM1, IM2, ... IMn (authenti cation) and, when the signature is verified, was not altered in transit (integrity).

As discussed above generating the plurality of the data pack ets DPI, DP2, ...DPn can also be based on a data provided by edge apps and related to the one or more industrial machines IM1, IM2, ... IMn. In this way the edge apps can generate new process relevant data, which can be further processed in form of data packets DPI, DP2, ...DPn.

Furthermore, the plurality of the data packets DPI, DP2, ...

DPn can be generated based on an infrastructure data, e.g. edge logfiles or on information about how frequently an edge device or an app on the edge had a malfunction or on billing information for pay-per-use of the edge apps etc. This infor mation can be of interest to the ecosystem provider and/or to edge app developer and/or to the manufacturer of the indus trial machines. Each signed data packet can then be forwarded to the data processing module DPM, which can e.g. read them (step S4) and compare their content with the content of the user-defined data filter DF, wherein the content of the user-defined data filter DF can comprise list(s) of words (whitelist(s)), matchmaking patterns or object descriptions. The data pro cessing module DPM can also apply matchmaking patterns of the data filter's DF whitelist to the content to get filtered da ta that matches the patterns.

If the data processing module DPM determines that there is no sensitive data and/or confidential information contained in the particular signed data packet, it lets the signed data packet through - arrow Y after step S5. In an embodiment the data processing module DPM can sign the signed data packet with a second digital signature PR2 and produce a double- signed data packet SDP1, SDP2, ... SDPn - step S6. This can be done to assure the integrity of the data, which passes the user-defined data filter DF. The signing can be performed by a hardware component, e.g. by a FPGA controller.

If, while comparing the content of a particular signed data packet with the user-defined data filter DF, the data pro cessing module DPM determines that the signed data packet contains sensitive data, it can reject it - arrow N after step S5. The term rejection can mean for example that the signed data packet is not forwarded further or that the data processing module DPM marks this data packet, e.g. by a mark "DENY" and processes it further, but only within the internal data processing system IDP, particularly only within the lo cal network of the industrial plant PL.

The first digital signature PR1 can be owned by the provider of the industrial machines IM1, IM2, ... IMn and/or by the pro vider of the industrial edge devices EDI, ED2, ...EDn and/or by the ecosystem provider. The second digital signature PR2 can be owned by the owner of the data processing module DPM and/or by the owner of the user-defined data filter DF (usu- ally by the industrial plant PL owner/operator). Both signa tures can be provided by a corresponding authority within the external data processing system EDP or within the internal data processing system IDP accordingly. In an embodiment both digital signatures PR1, PR2 are provided by an independent trusted entity, such as trust center TC illustrated in FIG 4. The trust center TC can be for example TUV.

All data packets that pass the user-defined data filter DF, e.g. the double-signed data packets SDP1, SDP2, ... SDPn can then be sent towards the external data processing system EDP - step S7.

In an embodiment signing with the first and/or second digital signature can be performed by means of a hardware component, e.g. by an FPGA controller.

In some embodiments the data processing module DPM can be a plugin installed on each industrial edge device EDI, ED2, ... EDn. The double-signed data packets SDP1, SDP2, ... SDPn can be, therefore, provided to the corresponding industrial edge device EDI, ED2, ...EDn for further processing, e.g. encryp tion. It will be appreciated by the person skilled in the art that the data processing module DPM can be designed as a net work component, which can be designed separately from the in dustrial edge device(s) EDI, ED2, ...EDn.

In an embodiment the data processing module DPM can encrypt the double-signed data packets.

The first digital signature PR1 of each data packet that passed the user-defined data filter DF, and can be for exam ple signed with the second digital signature PR2, i.e. the double-signed data packet SDP1, SDP2, ... SDPn, can be validat ed within the external data processing system EDP - S8, for example by the original equipment manufacturer (provider of the industrial machines) or by the provider of the industrial edge devices or by the provider of the software application that is carried out on the industrial edge devices. An ab sence of the data packets from the internal data processing system IDP can be noted by the external data processing sys tem EDP for control purposes. In this way it is possible for example to conclude that the user-defined data filter DF does not perform in a proper way. This can be due to an ill- defined content of the user-defined data filter DF, e.g. an ill-defined whitelist and/or matchmaking pattern and/or ob ject description, which can result in preventing all data from being uploaded into the external data processing system EDP, e.g. into the cloud.

An embodiment of the method includes storing of a first key PUB1 at the external data processing system EDP, wherein the first key PUB1 corresponds to the first signature PR1 and can be used to validate it (at any time in the future). This val idation can be performed for example by means of the edge backend device EB. FIG 4 illustrates that the first key PUB1 can be provided to the edge backend device EB.

As mentioned earlier, the internal data processing system IDP can further comprise the gateway component GW and all data packets that pass the user-defined data filter DF can be sent towards the external data processing EDP system through the at least one gateway component GW. FIG 3 and 4 show that these packets can be signed with the second digital signature PR2 before reaching the gateway GW.

The gateway component GW can be controlled by the industrial plant PL owner. If the data packets are signed by the data processing module DPM with the second digital signature PR2, the gateway component GW can be provided with a second key PUB2 for verifying the second digital signature PR2. In some embodiments the verification of the second digital signature PR2 can be performed somewhere in the internal data pro cessing system IDP but outside of the gateway component GW.

In an embodiment the verification of the second digital sig nature PR2 can be performed by an app within the internal network of the internal data processing system IDP. Such app can for example remotely access the gateway component GW in order to perform the verification. Verifying the second digi tal signature PR2, e.g. by the gateway component GW, increas es overall security of the scheme. In case of positive veri fication, the double-signed data packets SDP1, SDP2, ... SDPn can be sent to the external data processing system EDP, via an uplink UL, e.g. by the gateway component GW or by the app, which remotely accesses the gateway component GW for verify ing the second digital signature PR2. In an embodiment the double-signed data packets SDP1, SDP2, ... SDPn can be sent to the edge backend device EB.

In some embodiments the gateway component also can prevent the data packets marked with "DENY" from leaking outside of the internal network of the internal data processing system IDP, in particular, outside of the local network of the in dustrial plant PL.

In some embodiments the gateway component GW encrypts the da ta packets, for example the double-signed data packets, be fore sending them to the external data processing system EDP.

In an embodiment the encryption can be performed by an addi tional hardware component, e.g. FPGA controller.

As illustrated in FIG 4, in some embodiments the internal da ta processing system IDP can comprise a further network com ponent NC. This component can be located within the local network of the industrial plant PL, for example between the industrial edge devices EDI, ED2, ...EDn and the gateway com ponent GW. The network component NC can be a proxy-server.

The data processing module DPM, e.g. the plugin can be in stalled on the network component NC. In some embodiments the digital signatures (private keys) PR1, PR2 can be provided by a third trusted party, e.g. by a trust center TC. In this scenario the trust center TC can generate at least two key pairs KP1, KP2 for the provider of the industrial edge devic- es EDI, ED2, ...EDn and for the user (owner) of the industrial plant PL accordingly. The public keys PUB1, PUB2 for verifi cation of the digital signatures PR1, PR2 can be provided as well.

The above described embodiments of the present disclosure are presented for purposes of illustration and not of limitation. In particular, the embodiments described with regard to fig ures are only few examples of the embodiments described in the introductory part.

The reference signs in the claims used only for clarity pur poses and shall not be considered to be a limiting part of the claims.