WO2024012664A1

WO2024012664A1 - Cryptographically secure derived quality data of a metallic product

Info

Publication number: WO2024012664A1
Application number: PCT/EP2022/069486
Authority: WO
Inventors: Stefan GRÜLL; Thomas FÜRSTNER
Original assignee: S1Seven Gmbh
Priority date: 2022-07-12
Filing date: 2022-07-12
Publication date: 2024-01-18
Also published as: EP4555435A1

Abstract

Method and system for generating a cryptographically secure derived quality data of a metallic product, including data related to the origin and/or processing of the metallic product, the method comprising: performing a measurement of a process parameter during production or processing of the metallic product using a sensor (5), cryptographically attesting the measurement outcome (6) using a digital identity (7) associated with the sensor (5), submitting the attested measurement outcome (10) to a ledger database (3), receiving the attested measurement outcome (10) by a processor (4), computing a computation result (13) from the attested measurement outcome (10) and at least one computation instruction, digitally signing the computation result (13) using a digital identity (14) associated with the processor (4), submitting the signed computation result (15) to the ledger database (3).

Description

Cryptographically secure derived quality data of a metallic product

The present invention concerns a method for generating a cryptographically secure derived quality data of a metallic product , a use of such a method for generating a quality certi ficate for a metallic product , as well as a system comprising a production facility and configured for applying such a method . In this context , quality data of a metallic product includes data related to the origin, measured, or calculated physical and virtual properties and/or processing of the metallic product . The metallic product can be in particular a metal , metal ore or alloy .

Quality properties of important industrial raw materials such as metal vary depending on the batch . National and international standards speci fy the various types of quality certi ficates to be provided to the purchaser in accordance with the agreements made when the relevant products are ordered . These contain technical data on the product and the measurement results of the tests prescribed for the product in question . In all subsequent stages of further processing, these quality certi ficates are a central component of quality assurance . In many applications , especially safety-critical ones , there is an obligation to always keep the products and assemblies created traceable to the material batches used . Batch-related quality certi ficates are issued by an authori zed body, usually an accredited testing laboratory ( today via destructive testing) at the manufacturer ' s premises and enclosed with the physical delivery as an accompanying document . These quality certificates are part of the performance ful filment : steel without a quality certi ficate would be like a motor vehicle without registration papers .

In the future , quality data such as physical and mechanical properties can increasingly be derived from process data via calculation models , provided that the integrity of the data and the identity of the data source can be reliably confirmed . In addition virtual quality data such as C02 footprint and other sustainability attributes have to be computed based on process data collected during production and/or processing . Due to the document properties required by the regulator, such as stamp, signature or immutability of the data, quality documents described above are currently still created millions of times as paper or PDF documents and manually processed in the processing industry . Against the backdrop of advancing digiti zation, automation, and increasing documentation and compliance requirements , a development towards machine processability of documents is urgently needed .

The field and purpose of the present disclosure is to allow the required document properties to be generated for data obj ects as well , thus enabling the change from the exchange of documents to the exchange of veri fiable data for quality certi ficates .

Traditionally, raw and derived process data have been treated confidential . All or some information about a production and/or processing of a metallic product or a metallic article is generally inaccessible even to the owner of the product or article . Under these circumstances , trust in the properties and quality of the product or article had to be conveyed by testing . For example , quality certi ficates by trusted quality assurance authorities or test reports by trusted test facilities had to be taken into account and relied upon . This required a trusted exchange of material samples and material loss due to usually required destructive testing .

In principle , ledger databases such as blockchain-backed systems provide exceptional traceability and for this reason are a natural fit for quality control purposes . Such a system is described in US 2021 / 0278826 Al .

However, conventional blockchain-backed systems achieve trust by making public any and all information necessary to reproduce the content and therefore state of the blockchain . More speci fically, any derived data, usually referred to as "transactions" , receives trust by the fact that the derivation, including inputs and outputs , can be reproduced and checked by anyone participating in the blockchain . Under those circumstances , the steps necessary to perform the derivation must be also public .

It is an obj ect of the invention, to provide a traceable and trusted system and methodology based on a ledger database while at the same time allowing for confidentiality of the process data stored on the ledger database and/or of the computations used to generate derived quality data stored on the ledger database .

The disclosed method for generating a cryptographically secure derived quality data of a metallic product comprises : performing a measurement of a process parameter during production or processing of the metallic product using a sensor ; cryptographically attesting the measurement outcome using a digital identity associated with the sensor ; submitting the attested measurement outcome to a ledger database ; receiving the attested measurement outcome by a processor ; computing a computation result from the attested measurement outcome and at least one computation instruction; digitally signing the computation result using a digital identity associated with the processor ; and submitting the signed computation result to the ledger database .

The measurement can be, e . g . , pressure , temperature , weight , duration, flow rate , speed, power, voltage , position or distance , angle , force or generally any parameter related to the quality or performance of a metallic product or of a process for producing or processing a metallic product . Cryptographical attestation can be achieved by a digital signature of the measurement outcome and/or by encryption of the measurement outcome with a private signing and/or encryption key, respectively . Submission to the ledger database includes uploading the actual data, such as the attested measurement outcome or the signed computation result , to the ledger database or uploading the actual data to a storage service separate from the ledger database and uploading to the ledger database a cryptographic hash referencing the data uploaded to the storage service . The processor can be a processing entity or processing service provided separate from the sensor and/or the ledger database . Trust in the computation result is conveyed not through a structural relationship between the processor and the ledger database , but through the digital identity of the processor and its authentication .

The signed computation result may be submitted to the ledger database together with references to the attested measurement outcome in the ledger database .

The at least one computation instruction may comprise a sensor reference to the digital identity associated with the sensor, wherein receiving the measurement outcome includes querying the ledger database for the most recent signed measurement outcome submitted by the sensor referenced by the sensor reference . The computation might be triggered by the sensor . The sensor may be configured to send a push noti fication to the processor . Alternatively, the processor may be polling the ledger database for new data from the sensor .

After receiving the attested measurement outcome by a processor and before digitally signing the computation result the processor may veri fy the validity of the attestation of the attested measurement outcome . This may include veri fying a digital signature with a corresponding public key . Or decrypting an encrypted measurement outcome with an expected decryption key and checking for validity of the decrypted result .

The sensor may be a gauge calibrated and attested by a sensor certi fication authority, wherein the cryptographical attestation of the measurement outcome comprises a digital signature of the measurement outcome by the digital identity associated with the sensor, and wherein the digital identity associated with the sensor is certi fied by the sensor certi fication authority . The processor may store a whitelist of sensor certi fication authorities , wherein the processor veri fying the validity of the signature of the measurement outcome includes veri fying that the digital identity of the sensor is certi fied by a valid certi ficate issued by one of the certi fication authorities on the whitelist .

The attested measurement outcome may be an encrypted version of the measurement outcome provided by the sensor, wherein the encryption key is cryptographically associated with the digital identity associated with the sensor .

The attested measurement outcome may be obtained using a homomorphic encryption scheme , and wherein the at least one computation instruction is a homomorphic operation . The homomorphic encryption scheme may be a Paillier cryptosystem .

After receiving the attested measurement outcome by a processor and before computing the computation result , the processor may perform a filtering and selecting of the attested measurement outcome , wherein the filtering and selecting comprises decrypting the attested measurement outcome , and applying one or more predefined filtering criteria to the decrypted measurement outcome , wherein submitting the signed computation result computed from the attested measurement outcome is in response to the corresponding measurement outcome passing the one or more filtering criteria . The filtering may include selecting or rej ecting a measurement outcome or computation result according to certain predefined criteria . When the criteria are not met , the instance can be rej ected and the processing aborted, such that nothing is submitted to the ledger database .

The method may comprise retrieving the signed computation result from the ledger database , receiving an unblinding key which corresponds to the encryption key cryptographically associated with the digital identity associated with the sensor, and decrypting the computation result using the unblinding key to obtain an unblinded computation result . The at least one computation instruction may comprise a type constraint defining an acceptable measurement type of the measurement outcome , wherein the certi ficate of the digital identity associated with the sensor defines the measurement type for which the gauge is calibrated and attested, wherein before computing the computation result the processor veri fies that the type constraint is respected .

The processor may be implemented in a trusted execution environment . The trusted execution environment ( TEE ) protects the executed instructions (program or code ) and the processed data with respect to confidentiality and integrity . It prevents unauthori zed alteration of the processed data before , during and after processing by any entity outside the TEE .

The at least one computation instruction may be locked inside the trusted execution environment . In this case , the TEE ensures that the instructions in the TEE cannot be replaced or modi fied by unauthori zed entities .

Before computing the computation result the processor may receive the at least one computation instruction, wherein the at least one computation instruction is digitally signed, and the processor veri fies the validity of the signature of the at least one computation instruction .

The processor may veri fy the validity of the signature of the at least one computation instruction includes veri fying that the digital identity associated with the signature is certi fied by a valid certi ficate issued by a certi fication authority on a whitelist of instruction certi fication authorities stored by the processor .

The attested measurement outcome may be submitted to the ledger database together with one or more measurement parameters selected from a group comprising : timestamp, geographic location, ambient temperature , ambient pressure , error message of the sensor, warning message of the sensor . The method disclosed above may be used for generating a quality certi ficate for a metallic product , in particular in the production of steel . This use has the advantage that otherwise necessary destructive testing can be avoided .

In this use , the signed computation result may be submitted to the ledger database together with a product identi fier .

The submission of the signed computation result and the product identi fier may have the form of a transaction to a digital identity of a producer of the product .

The disclosure also concerns a method for initiali zing a system, for example a process for the setup of the data generating entities , such as the sensors and processors , including their attestation .

Finally, the disclosure concerns a system comprising a production facility having at least one sensor and at least one processor, wherein the production facility is configured for producing or processing a metallic product , wherein the sensor is configured for performing a measurement of a process parameter of the production process of the metallic product or of the processing of the metallic product and cryptographically attesting the measurement outcome using a digital identity associated with the sensor and submitting the attested measurement outcome to a ledger database , wherein the processor is configured for receiving the attested measurement outcome and computing a computation result from the attested measurement outcome and at least one computation instruction and digitally signing the computation result using a digital identity associated with the processor and submitting the signed computation result to the ledger database .

Referring now to the drawings , wherein the figures are for purposes of illustrating the present invention and not for purposes of limiting the same ,

Fig . 1 schematically shows a first embodiment of the present disclosure , wherein a computation result is computed from a single measurement outcome and submitted to a ledger database ;

Fig . 2 schematically shows a second embodiment of the present disclosure , wherein a computation result is computed from two measurement outcomes and using computation instructions retrieved from the ledger database ;

Fig . 3 schematically illustrates the identities and references within the ledger database according to Fig . 2 ;

Fig . 4 and 5 schematically illustrate the registration of identities on the blockchain of a sensor and of an instruction certi fication authority; and

Fig . 6 schematically shows a third embodiment of the present disclosure , wherein the measurement outcome is encrypted and the computation operates on the encrypted measurement outcome .

Fig . 1 shows a system 1 comprising a production facility 2 for producing or processing a metallic product , a ledger database 3 and a processor 4 . The production facility 2 has at least one sensor 5 . In Fig . 1 the production facility 2 is schematically shown with only a single sensor 5 for the sake of simplicity . The production facility 2 performs a production process . The sensor 5 measures a process parameter during the production process . From the measurement the sensor obtains a measurement outcome 6 indicated by a sheet with a scale . The sensor 5 cryptographically attests the measurement outcome 6 using a digital identity 7 . The digital identity 7 is indicated by a key . Speci fically, the key is a private key stored in a secure element 8 integrated with or securely attached to the sensor 5 . The secure element 8 comprises a cryptoprocessor 9 for securely accessing and using the digital identity 8 . Either way, the digital identity 7 used by the sensor 5 is physically and logically associated with the sensor 5 . The sensor 5 submits the attested measurement outcome 10 to the ledger database 3 . The attested measurement outcome 10 is indicated by a sheet with the scale and a seal 11 . The seal 11 represents a digital signature of the measurement outcome 6 . In this embodiment , the measurement outcome 6 is stored in the ledger database 3 in a publicly readable format . This is useful for attesting process data where the value for third parties lies in the metallic product and not in the process data itsel f . The process data serves as a quality assurance for the metallic product .

In this embodiment , the processor 4 uses an embedded program 12 . The embedded program 12 comprises a collection of computation instructions , for example a particular sequence of computation instructions to be performed in order to obtain a computation result . At least one of the computation instructions comprises a sensor reference . The sensor reference is to the digital identity 7 associated with the sensor 5 .

Before performing the computation instruction using the measurement outcome 6 , the processor 4 receives the attested measurement outcome 10 . For that purpose , the processor 4 queries the ledger database 3 for the most recent attested measurement outcome 10 submitted by the sensor 5 referenced by the sensor reference . The ledger database 3 responds to the processor 4 by providing and transmitting the most recent suitable ( signed and attested) measurement outcome 10 . The processor 4 then computes a computation result 13 from the attested measurement outcome 10 by applying and carrying out the one or more instructions of the program 12 to the measurement outcome 6 . The obtained computation result 13 is indicated by a sheet with a formula sign " fx" .

The computation instructions for example contain information for trans forming the measurement outcome 6 into derived quality data provided as the computation result 13 . Such a trans formation may employ know-how about the process of the production facility 2 , which can be embedded in the computation instructions . For example , there might be known ( static or constant ) properties of that process , which permit to trans form the measurement outcome 6 into a more meaningful computation result 13 . The processor 4 digitally signs the computation result 13 using a digital identity 14 . Again, the digital identity 14 is indicated by a key, which is now a dif ferent key than that of the sensor 5 . This digital identity 14 is associated with the processor 4 ( or, optionally with the program 12 ) . The processor 4 submits the signed computation result 15 to the ledger database 3 .

In this example , the measurement outcome 6 and the computation result 13 are submitted to the same ledger database 3 . This is not necessary, but advantageous in order to provide a traceable history of computation results and of the measurement outcomes they are based on within the same ledger database . A single cohesive ledger database guarantees meaningful and reproducible relationships and sequences between measurement outcomes and derived computation results , i . e . , the derived quality data of the metallic product .

Fig . 1 thus illustrates a simple embodiment of the disclosed method for generating a cryptographically secure derived quality data . The derived quality data is the computation result 13 . It is cryptographically secure because it is obtained from a cryptographically verified input , the attested measurement outcome 10 , and represented by a cryptographically attested output , the digitally signed computation result 15 . For that purpose , the sensor 5 and the processor 4 are both associated with their own digital identity 7 , 14 indicated by two di f ferent keys . Both of their keys are managed in a secure element 8 , 16 embedded with the sensor 5 and the processor 4 respectively . For obtaining this cryptographically secure derived process data, the following steps are carried out :

The sensor 5 performs a measurement of a process parameter during production or processing of the metallic product . The process parameter can be any physical parameter related to the supervised process . In the case of a production process , it can for example be a pressure value of a pressured fluid connection or of a pressure tank . This is indicated schematically by a pressure gauge . The sensor 5 employs its secure element 8 to cryptographically attest the measurement outcome 6 using the digital identity 7 . The sensor 5 then submits the attested measurement outcome 10 to the ledger database 3 . The attested measurement outcome 10 is submitted to the ledger database 3 together with one or more measurement parameters . The measurement parameters submitted together with the attested measurement outcome 10 are selected from a group comprising : timestamp, geographic location, ambient temperature , ambient pressure , error message of the sensor, warning message of the sensor . In the present example , the pressure value forming the measurement outcome 6 is submitted together with a timestamp of the measurement and any available error message of warning message of the sensor 5 . These measurement parameters are attested together with the measurement outcome 6 . That is , the secure element 8 generates a digital signature over a measurement record comprising the measurement outcome 6 and the measurement parameters . The complete record and digital signature are then submitted to the ledger database 3 .

The processor 4 runs the embedded program 12 . The program 12 asks for an input in the form of a measurement outcome 6 of an identi fied sensor . The processor 4 therefore queries the ledger database 3 and fetches the attested measurement outcome 10 previously submitted by the identi fied sensor 5 . The processor 4 veri fies the validity of the attestation of the attested measurement outcome 10 .

From the attested measurement outcome 10 and the computation instructions of the program 12 , the processor 4 computes the computation result 13 . The processor 4 is implemented in a trusted execution environment ( TEE ) . The TEE is a speciali zed hardware component within a general-purpose central processing unit ( CPU) . The at least one computation instruction is locked inside the trusted execution environment . The trusted execution environment guarantees the content ( authenticity and integrity) of the computation instructions carried out by the processor 4 . I f the computation within the trusted execution environment is success ful and the validity of the attestation of the attested measurement outcome 10 provided as an input to the trusted execution environment has been positively veri fied, the processor 4 then employs its secure element 16 to digitally sign the computation result 13 using its digital identity 14 . Finally, the processor 4 submits the signed computation 15 result back to the ledger database 3 .

A second, more extensive example is illustrated in Fig . 2 . In addition to the production facility 2 , the ledger database 3 and the processor 4 , the fourth participant is the instruction provider 17 . The instruction provider 17 is schematically indicated by a microscope . For example , a person such as a researcher 18 comes up with a program 19 comprising the one or more instructions for trans forming a measurement outcome into derived process data . The program 19 is indicated by a sheet with a gear wheel . The instruction provider 17 has its own digital identity 20 indicated by yet another key managed in a secure element 21 . In order to contribute their program 19 in the present framework, the instruction provider 17 digitally signs their newly developed program 19 with their digital identity 20 . Then they submit the digitally signed program 22 to the ledger database 3 for anyone to veri fy and use .

The production facility 2 in this example has not only one , but two sensors 5 , 23 . Each sensor 5 , 23 has its own associated digital identity 7 , 24 . They both submit measurement outcomes 10 , 25 attested by the respective digital identity 7 , 24 to the ledger database 3 . Over time , the ledger database 3 stores measurement outcomes from di f ferent times of the production process . Previous attested measurement outcomes 26 , 27 are maintained in the history of the ledger database 3 . The processor 4 when running the program 19 accesses only the most recent of each type of measurement outcome .

In this example , the processor 4 does not rely purely on an embedded program, but before computing the computation result 28 the processor 4 queries the ledger database 3 for an updated program and receives the most recent program 22 with the at least one computation instruction . As indicated above , the at least one computation instruction is digitally signed by the instruction provider 17 . The processor 4 veri fies the validity of the signature of the at least one computation instruction . For this purpose , as a first step, the processor determines the digital identity 20 associated with the signature , i . e . , of the instruction provider 17 having generated the digital signature of the computation instructions . The processor veri fies whether that digital identity 20 of the instruction provider is certi fied by a valid certi ficate 29 ( see Fig . 5 ) . A certi ficate 29 is valid for this purpose when it is issued by a certi fication authority, wherein the digital identity 30 of the certi fication authority is on a whitelist of instruction certi fication authorities . The processor 4 stores such a whitelist in an embedded secure memory for the purpose of performing these kinds of veri fications . That way, the authenticity of the computation instructions employed by the processor 4 can be relied upon . In other words , the content of the computation instruction is attested by the instruction provider 17 , whose identity 20 in turn is attested by an accepted (by way of whitelisting) instruction certi fication authority . The processor 4 is merely required to attest to the accurate execution of those computation instructions , which is achieved by the TEE .

The program 19 of the example shown in fig . 2 uses two measurement outcomes of di f ferent types . In other words , they concern di f ferent process parameters . The first required measurement outcome is a pressure value , similar to the example in fig . 1 , provided by a pressure gauge acting as a first sensor 5 . The second required measurement outcome is a weight value provided by a scale acting as a second sensor 23 . In order to avoid errors in creating sensor relationships suitable for supplying parameters to a given program, the at least one computation instruction comprises type constraints defining acceptable measurement types of the measurement outcomes to be used together with the program 19 . In this example , the program 19 may define that a first input parameter 31 must be a pressure value and a second input parameter 32 must be a weight value . The certi ficate of the digital identity associated with each of the sensors 5 , 23 defines the measurement type for which the gauge or scale are calibrated and attested respectively . The digital identity 7 of the pressure gauge can only be validly used to attest pressure measurement outcomes and the digital identity 24 of the scale can only be validly used to attest weight measurement outcomes . Before computing the computation result 28 the processor 4 veri fies that the type constraints are respected, i . e . , the computation uses only values attested from a digital identity valid for providing pressure measurements for the first required measurement outcome and only values attested from a digital identity valid for providing weight measurements for the second required measurement outcome . The signed computation result 33 is submitted to the ledger database 3 together with references to the attested measurement outcomes 10 , 25 in the ledger database 3 .

Fig . 3 illustrates the identities and attestations and references within the ledger database 3 more in detail . The ledger database 3 is illustrated as a chronological list of entries 34 , starting at the top 35 and adding new entries in the vertical direction 36 at the bottom 37 . The most recent entry 38 stored in the ledger database 3 shown at the very bottom is the signed computation result 33 . The signature of the signed computation result 33 refers to the digital identity 39 of the processor 4 . The entry of the computation result comprises a reference 40 to another entry 41 containing the program 22 . The program 22 defines the at least one instruction that have been performed in order to obtain the computation result 33 . The program entry 41 in the ledger database 3 is shown at the very top . The signature 42 of the program 22 refers to the digital identity 20 of the instruction provider 17 . The program 22 itsel f defines the type of two input parameters 31 , 32 . Based on these type references , the most recent attested measurement outcomes 10 , 25 stored in the ledger database 3 and having a suitable type are referenced from the program 22 . In this simpli fied example , the type references can be directly mapped to sensor references 43 , 44 , because there is only one suitable sensor for each type of measurement outcome . Alternatively, the processor 4 may include direct references with the signed computation result 33 that point to the particular measurement outcomes that have been used in obtaining it . The signature 45 of the first measurement outcome 10 refers to the digital identity 7 of the first sensor 5 . The signature 46 of the second measurement outcome 25 refers to the digital identity 24 of the second sensor 23 .

Fig . 4 and 5 further illustrate the certi fication path and registration of sensors and processes , which can also be documented in a traceable manner in the ledger database 3 . Generally, while the ledger database 3 is indicated with di fferent entries 34 in Fig . 2 , 3 , 4 and 5 , all those entries 34 can be part of the same ledger database 3 . Each drawing indicates only a selection of entries necessary for describing di fferent aspects of the present disclosure . These lists of entries 34 are non-exclusive . For example , the first sensor 5 is a gauge calibrated and attested by a sensor certi fication authority . The cryptographical attestation of the measurement outcome 10 comprises a digital signature 45 of the measurement outcome by the digital identity 7 associated with the sensor 5 . The digital identity 7 associated with the first sensor 5 is certi fied by the sensor certi fication authority . This is accomplished by including a registration record 47 in the ledger database 3 , which comprises a reference 48 to the digital identity 7 of the sensor 5 . The registration record 47 is digitally signed with the digital identity 49 of the sensor certi fication authority . The processor 4 stores a whitelist of sensor certi fication authorities . The processor 4 veri fying the validity of the signature 45 of the measurement outcome 10 includes veri fying that the digital identity 7 of the sensor is certi fied by a valid certi ficate issued by one of the certi fication authorities on the whitelist . Similarly, the digital identity 20 of the instruction provider 17 can be attested by a registration record 50 stored in the ledger database 3 , wherein this registration record 50 comprises a reference 51 to the digital identity 20 of the instruction provider 17 and is digitally signed with the digital identity 30 of an instruction provider certi fication authority . The preferred use-case of the present disclosure is for generating a quality certi ficate for a metallic product . For the purpose of providing a quality certi ficate or generally providing information allowing for a quality audit of metallic product , the signed computation result 33 can be submitted to the ledger database 3 together with a product identi fier . The product identi fier can be used to link the information stored in the ledger database 3 to the actual physical metallic product or metallic article . For example , the product identi fier may be attached to the metallic product or metallic article itsel f when it leaves the production or processing process . In another example , the product identi fier may be linked to a batch of metallic products and/or to a particular time stamp when a metallic product has been completed or delivered . The product identi fier may also be associated with the metallic product via intrinsic physical properties of the metallic product , for example unique or suf ficiently random properties that can be determined relatively easily, preferably non-destructively, from the metallic product . There may be a reference table ( on the ledger database or external to it ) storing associations between such intrinsic properties and corresponding product identi fiers .

Speci fically, the signed computation result 33 and the product identi fier can be submitted to the ledger database 3 as a transaction . The recipient of the transaction is a producer or processor of the metallic product . This recipient is identi fied by their digital identity .

The third embodiment , which is illustrated in Fig . 6 , is a variation of the first embodiment shown in Fig . 1 . In this embodiment , the attested measurement outcome 52 is an encrypted version of the measurement outcome 6 provided by the sensor 5 . The encryption key can be cryptographically associated with the digital identity 7 associated with the sensor 5 . Alternatively, the encryption key can be a separate key, for example , a shared secret 53 embedded in the secure element of the sensor 5 and unknown to the processor 4 . In the present embodiment , the attested measurement outcome 10 is obtained using a homomorphic encryption scheme . Homomorphic encryption is a known technique and there are several cryptosystems available , depending on the type of operations that need to be performed . The most general ( in fact arbitrary) operations are possible in fully homomorphic encryption ( FHE ) cryptosystems . There are several open-source FHE libraries available implementing di f ferent FHE schemes , such as HElib, PALISADE or Lattigo . Because the processor 4 does not have the decryption key, the at least one computation instruction of the program 54 in this embodiment needs to be a homomorphic operation compatible with the employed homomorphic encryption cryptosystem . Consequently, the computation result 55 also remains encrypted, as well as the signed computation result 56 that the processor 4 submits to the ledger database 3 .

Claims

Claims :

1. Method for generating a cryptographically secure derived quality data of a metallic product, including data related to the origin and/or processing of the metallic product, the method comprising : performing a measurement of a process parameter during production or processing of the metallic product using a sensor (5) , cryptographically attesting the measurement outcome (6) using a digital identity (7) associated with the sensor (5) , submitting the attested measurement outcome (10) to a ledger database ( 3 ) , receiving the attested measurement outcome (10) by a processor ( 4 ) , computing a computation result (13) from the attested measurement outcome (10) and at least one computation instruction, digitally signing the computation result (13) using a digital identity (14) associated with the processor (4) , submitting the signed computation result (15) to the ledger database (3) .

2. Method according to claim 1, characterised in that the signed computation result (15) is submitted to the ledger database (3) together with references to the attested measurement outcome in the ledger database (3) .

3. Method according to claim 1 or 2, characterised in that the at least one computation instruction comprises a sensor reference (44) to the digital identity (7) associated with the sensor (5) , wherein receiving the measurement outcome (6) includes querying the ledger database (3) for the most recent signed measurement outcome submitted by the sensor (5) referenced by the sensor reference (44) .

4. Method according to any one of claims 1 to 3, characterised in that after receiving the attested measurement outcome (10) by a processor (4) and before digitally signing the computation result (13) the processor (4) verifies the validity of the attestation of the attested measurement outcome (10) .

5. Method according to any one of claims 1 to 4, characterised in that the sensor (5) is a gauge calibrated and attested by a sensor certification authority, wherein the cryptographical attestation of the measurement outcome (6) comprises a digital signature (45) of the measurement outcome (6) by the digital identity (7) associated with the sensor (5) , and wherein the digital identity (7) associated with the sensor (5) is certified by the sensor certification authority.

6. Method according to claims 4 and 5, characterised in that the processor (4) stores a whitelist of sensor certification authorities, wherein the processor (4) verifying the validity of the signature (45) of the measurement outcome (6) includes verifying that the digital identity (7) of the sensor (5) is certified by a valid certificate issued by one of the certification authorities on the whitelist.

7. Method according to any one of claims 1 to 6, characterised in that the attested measurement outcome (10) is an encrypted version of the measurement outcome (6) provided by the sensor (5) , wherein the encryption key is cryptographically associated with the digital identity (7) associated with the sensor (5) .

8. Method according to claim 7, characterized in that the attested measurement outcome (10) is obtained using a homomorphic encryption scheme, and wherein the at least one computation instruction is a homomorphic operation.

9. Method according to claim 7 or 8, characterized in that after receiving the attested measurement outcome (10) by a processor (4) and before computing the computation result (13) , the processor (4) performs a filtering and selecting of the attested measurement outcome (10) , wherein the filtering and selecting comprises decrypting the attested measurement outcome (10) , and applying one or more predefined filtering criteria to the decrypted measurement outcome, wherein submitting the signed computation result (15) computed from the attested measurement outcome (10) is in response to the corresponding measurement outcome passing the one or more filtering criteria.

10. Method according to any one of claims 7 to 9, characterised in that it comprises retrieving the signed computation result (15) from the ledger database (3) , receiving an unblinding key which corresponds to the encryption key cryptographically associated with the digital identity (7) associated with the sensor (5) , and decrypting the computation result (13) using the unblinding key to obtain an unblinded computation result.

11. Method according to claim 6, characterised in that the at least one computation instruction comprises a type constraint defining an acceptable measurement type of the measurement outcome (6) , wherein the certificate of the digital identity (7) associated with the sensor (5) defines the measurement type for which the gauge is calibrated and attested, wherein before computing the computation result (13) the processor (4) verifies that the type constraint is respected.

12. Method according to any one of claims 1 to 11, characterised in that the processor (4) is implemented in a trusted execution environment .

13. Method according to claim 12, characterised in that the at least one computation instruction is locked inside the trusted execution environment.

14. Method according to any one of claims 1 to 13, characterised in that before computing the computation result (13) the processor (4) receives the at least one computation instruction, wherein the at least one computation instruction is digitally signed, and the processor (4) verifies the validity of the signature of the at least one computation instruction.

15. Method according to claim 14, characterised in that the processor (4) verifying the validity of the signature of the at least one computation instruction includes verifying that the digital identity (20) associated with the signature is certified by a valid certificate issued by a certification authority on a whitelist of instruction certification authorities stored by the processor ( 4 ) .

16. Method according to any one of claims 1 to 15, characterised in that the attested measurement outcome (10) is submitted to the ledger database (3) together with one or more measurement parameters selected from a group comprising: timestamp, geographic location, ambient temperature, ambient pressure, error message of the sensor, warning message of the sensor.

17. Use of the method according to any one of claims 1 to 16 for generating a quality certificate for a metallic product.

18. Use according to claim 17, characterised in that the signed computation result (15) is submitted to the ledger database (3) together with a product identifier.

19. Use according to claim 18, characterised in that the submission of the signed computation result (15) and the product identifier has the form of a transaction to a digital identity of a producer of the product.

20. System comprising a production facility (2) having at least one sensor (5) and at least one processor (4) , wherein the production facility (2) is configured for producing or processing a metallic product, wherein the sensor (5) is configured for performing a measurement of a process parameter of the production process of the metallic product or of the processing of the metallic product and cryptographically attesting the measurement outcome (6) using a digital identity (7) associated with the sensor (5) and submitting the attested measurement outcome (10) to a ledger database (3) , wherein the processor (4) is configured for receiving the attested measurement outcome (10) and computing a computation result (13) from the attested measurement outcome (10) and at least one computation instruction and digitally signing the computation result (13) using a digital identity (14) associated with the processor (4) and submitting the signed computation result (15) to the ledger database (3) .