WO2016189084A1

WO2016189084A1 - Social media and industrial internet of things

Info

Publication number: WO2016189084A1
Application number: PCT/EP2016/061897
Authority: WO
Inventors: Simo Säynevirta
Original assignee: Abb Schweiz Ag
Priority date: 2015-05-26
Filing date: 2016-05-26
Publication date: 2016-12-01

Abstract

To facilitate information exchange between the Industrial Internet of Things and the social media service dedicated to human interactions, an apparatus running an analytic tool analysing data comprising data generated by the Industrial Internet of Things, is configured to detect a trigger event for the social media service, and then determine one or more first pointers pointing to one or more analysed data sets analysed by the analytic tool and a second pointer pointing to analysis information indicating at least the analytic tool, further determine one or more key-words, convert the one or more first pointers, the second pointer and the one or more keywords to a post in a format used in the social media service, and send the post to a social media server providing the social media service.

Description

SOCIAL MEDIA AND INDUSTRIAL INTERNET OF THINGS

FIELD

The present invention relates to social media and Industrial Internet of

Things. BACKGROUND ART

The evolvement of networking between computers and computing devices used by multiple people at disparate locations eventually has led to the development of online social activities that involve multiple people, typically at more than one physical location, with interactions mediated via one or more networks. These online social activities are available via different social media services. In a technical point of view, a social media service is a digital platform via which individuals can create and share user-generated content. Typical examples of social media services are Internet-based micro-blogging services, like Twitter and Face- book. The basic principle is that a user joins a social media service by opening an account, and after that can publish data and subscribe to data published by other users. A user may ask some advice or post a thought, and the other users may provide advice or otherwise comment the published data so that the advice/comment is linked to the published data. The same way of collaboration has been adopted for professional use as well. Enterprise social media services, like Yammer, are pri- vate social media services with restricted access that provide an internal platform to employees in an organization, or to a group of employees, for example, to exchange information, knowledge, documents, task management, contact data, etc. across departments and locations. Today enterprise social media services are increasingly used as a source of professional information, or a starting point to gather information or to find a solution to an open question by posting the question to the network.

The evolvement of networking between computers and computing devices, especially different sensors, capable to communicate over the Internet without user involvement, has also led to a so called Internet of Things intended for domestic appliances, and to Industrial Internet of Things intended for professional/enterprise use. The Industrial Internet of Things is a network concept which connects an industrial Thing and its embedded one or more sensors, for example, with the Internet, for open information exchange and communication, in order to achieve tracking, monitoring and management of the Thing, for example. In the Industrial Internet of Things, the "Things" are individual devices, equipment, systems, sub-systems, or processes in different industrial environments. For example, the Industrial Internet of Things enable remote control, optimization at the level of the entire system, and use of sophisticated machine-learning algorithms that take into account vast quantities of data generated by Things connected to the Industrial Internet of Things as well as the external context of every individual Thing. In the Industrial Internet of Things, human individuals, such as operators of an industrial plant, are typically not considered as "things" but as users of the information who can react to alerts created by the system, analyse the data and take a decision based on the insight gained. In other words, information exchange in the Industrial Internet of Things is a machine-to-machine based information exchange system.

There are machine-to-machine based system applying concepts and technologies of social media services. For example, US 8434095 describes a solu- tion in which enterprise application events are encoded, by a software client, separate from the enterprise application, in a micro-blog compatible format to create an event post, which is then uploaded to a micro-blog server. Information technology (IT) professionals then receive the information, and use the event information as an aid in maintaining, repairing, or understanding the operation of the enter- prise application software. Another example of a machine-to-machine based system applying concepts and technologies of social media services is Dweet.io that is a Twitter-like social media service dedicated to devices, in which devices publish some parameter values, for example a public pool posting its temperature every hour, and human individuals may follow or simply view the posts. Still a further example of a kind of a social media service dedicated to devices is disclosed in US2014/0244768 that teaches how home appliance IoT devices can employ common social networking capabilities to interact with other home appliance IoT devices. Yet another example is disclosed in US 2014/0244770 in which a social media service server interacts, upon a user input to obtain certain measurements re- suits, such as a temperature, from a specific location, with a machine-to-machine service server as if the social media service server were a machine-to-machine device.

While studying the alerts or other analysis information, especially in tricky situations in which the operator, or an IT professional, is not aware of a straightforward solution to the problem, he/she could obtain some help, like solution proposals, and/or some valuable information from his colleagues using the user-to-user social media service, but a machine-to-machine based system, like the Industrial Internet of Things, and the user-to-user based social media service do not collaborate with each other. Therefore a user wanting to use information in both systems is posed with extra work in describing the context of the issue in the Industrial Internet of Things, for example, to the social media service.

SUMMARY

An object of the present invention is to facilitate information exchange between the Industrial Internet of Things and the social media service dedicated to human interactions. The object of the invention is achieved by a method, an apparatus, a computer program product and a system which are characterized by what is stated in the independent claims. The preferred embodiments of the invention are disclosed in the dependent claims.

A general aspect of the invention uses posts that are sent from an anal- ysis system to be published in a social media service dedicated to human interactions, a post comprising at least system generated information enabling access to data used in an analysis relating to the post.

BRIEF DESCRIPTION OF THE DRAWINGS

In the following, exemplary embodiments will be described in greater detail with reference to accompanying drawings, in which

Figure 1 shows simplified architecture of a system and block diagrams of some apparatuses according to an exemplary embodiment;

Figure 2 depicts exemplary information exchange;

Figures 3 and 4 are flow charts illustrating exemplary functionalities; Figures 5A to 5G are block diagrams illustrating exemplary use cases; and

Figure 6 is a block diagram of an exemplary apparatus.

DETAILED DESCRIPTION OF SOME EMBODIMENTS

The following embodiments are exemplary. Although the specification may refer to "an", "one", or "some" embodiment(s) in several locations, this does not necessarily mean that each such reference is to the same embodiment(s), or that the feature only applies to a single embodiment. Single features of different embodiments may also be combined to provide other embodiments. The present invention is applicable to any automated analysis system that is configured to analyse data generated by sensors, machines and other devices that are part of an Industrial Internet of Things, or a corresponding industrial system generating data. It should be appreciated that the automated analysis sys- tern may be a semi-automated system requiring human assistance in the analysis, or a fully automated system not using any human assistance in the analysis.

A general exemplary architecture of a system is illustrated in Figure 1. Figure 1 is a simplified system architecture only showing some devices, apparatuses and functional entities, all being logical units whose implementation may dif- fer from what is shown. The connections shown in Figure 1 are logical connections; the actual physical connections may be different. It is apparent to a person skilled in the art that the systems also comprise other apparatuses, devices, functions and structures used in or for big data, data management, and communication in the system or in one part of the system. They, as well as the protocols used, are well known by persons skilled in the art and are irrelevant to the actual invention. Therefore, they need not to be discussed in more detail here.

In the example illustrated in Figure 1, the system 100 comprises an automated system 101 connected over a network 102 to a social media server 120 providing a social media service platform for a social media service dedicated to human interaction. The system 100 may be an enterprise system, or a combination of one or more enterprise-systems and/or one or more "non-enterprise" systems. Although the automated system 101 may provide social media services dedicated to thing/device interaction, below the social media service means a human social media network/service, or a social media service dedicated to human interaction. Further, a post means herein any content that is to be published in a social media service.

The automated system 101 comprises the Industrial Internet of Things 103, and an analysis system depicted by an analyser apparatus 110. For example, in a multi-enterprise scenario the system 100 may comprise a maintenance service providing maintenance services to different enterprises as an analysis system or part of the analysis system, and each enterprise using the maintenance services has its own Industrial Internet of Things.

As said above, the Industrial Internet of Things 103 comprises as Things different devices, machines, apparatuses, equipment, systems, sub-systems, pro- cesses etc., There are no restrictions what constitutes a Thing, it suffices that the Thing comprises means for performing one or more different measurements on environment and/or one or more operations, for example, and means for sending the information at least to the analyser apparatus 110. Examples of Things are depicted in Figure 1 by a sensor 130, a device 130' and a sub-system 130", that itself comprises Things (not illustrated in Figure 1) that the analysis system sees as one, combined Thing, the sub-system 130". The implementation of the Industrial Internet of Things, data collected therefrom and means used for information exchange bears no significance to the invention, and therefore they are not described in more detail here. It is obvious for one skilled in the art that any known or future solution may be used.

The analyser apparatus 110, although depicted as one apparatus, may be a distributed apparatus. In order to utilize data generated by the Things in the Industrial Internet of Things, the analyser apparatus 110 comprises a data storing unit 113 configured to store the data 112-1 generated by the Things to a memory 112, and one or more analytic tools 114 with which the data 112-1, or sub-sets of the data 112-1, may be analysed. The data 112-1 may also comprise links to social media service discussions, or copies of the discussions, and/or copies of keywords, and thereby form an enhanced knowledge base combining the two different information sources. A knowledge base represents facts about the world, i.e. it stores structured and unstructured information, which are typical complex information.

An analytic tool is a special purpose analysis software, or a software suite (application suite) that is a collection of software of related functionality, often sharing a more-or-less common user interface and some ability to smoothly exchange data with each other. The analytic tool may be based on data visualization, data mining, mathematical models of an industrial process used, machine learning, such as deep learning, unsupervised learning, semi-supervised learning, supervised learning, anomaly detection, and self-learning artificial intelligence, etc. However, the actual functionality and the purpose of the analytic tool are not relevant for the invention and therefore the analytic tool is not described in more detail here. It is obvious for one skilled in the art that any known or future analytic tool may be used. For example, an analytic tool may be created for scanning and/or monitoring and/or analysing the enhanced knowledge base, or the social media service discussions part of the enhanced knowledge base.

To support interaction with the social media server 120, and hence with the social media service, the analyser apparatus 110 comprises a social media unit 115, and the memory comprises data model definitions 112-2, and at least two kinds of lists: a list 112-4 for analysis information, and a list 112-3 for Industrial Internet of Things context and the corresponding measured data. Exemplary functionalities of the social media unit are described in more detail below.

The data model definitions 112-2 may be according to ISA-95 type equipment model definitions, for example. ISA-95 is an ISO standard defining, among other things, standard terminology and information models, which can be used to decide which information should be exchanged between enterprise systems and control systems. In addition to ISA-95 type equipment model definitions, the data model definitions may comprise a timestamp of the measurement, value itself, its quality attributes, etc. It should be appreciated that any other data model definitions may be used as well, as long as there are one or more semantic definitions for the data so that the raw data can be indexed.

The list 112-3 for the Industrial Internet of Things context and the corresponding measured data associates received raw data with its indexed context (data context) and contains one or more pointers (Pointer 1) to the memory area the received raw data, and possibly the data context, is stored. The list 112-3 for the Industrial Internet of Things context may further comprise a link of the post and/or information on used keywords, for example a list of used keywords, and/or one or more pointers to one or more predefined keyword sets stored for example in association with a data model definitions, or data sets.

The list 112-4 for the analysis information is for social media posts, and contains a pointer at least to an analytic tool used when a post has been created. The list 112-4 for the analysis information may further associate the analytic tool with its configuration. The configuration may contain used search criteria and parameters used by the analytic tool in its analytic method, for example. The list 112- 4 for the analysis information may further comprise a link of the post and/or information on used keywords, for example a list of used keywords, and/or one or more pointers to the one or more predefined keyword sets stored for example in association with the data model definitions, or data sets. Further, the list 112-4 may comprise a link to analysis results, if the analysis tool is configured to provide recursive analysis and store analysis results to the memory. The analysis results, for example Fourier transforms of vibration results, such as Fourier transforms (spectra) calculated from high resolution raw data collected from vibration sensors, may then form a new data that is usable as a starting point for a further analysis.

In the illustrated example, the analyser apparatus 110 comprises fur- ther a user interface 111 for human-machine interaction. Users may study the data, start running an analytic tool, view outputs/results, provide different inputs, etc. via the user interface 111. Examples of the user interface include standard input devices, such as a keyboard, motion detection device, mouse, scanner and microphone, standard output devices, such as a display, screen, loudspeakers and printers, different kinds of headsets, such as smart glasses and virtual helmets, and mul- timodal devices, such as a wired glove or omnidirectional intelligent clothing. In other words, any kind of a user interface, including future ones, may be used.

As to the internal structure of the analyser apparatus, it should be appreciated that the data storing unit, or part of its functionality, and the social media unit may be integrated together. Further, although herein the analytic tool func- tionality and the social media unit functionality are not integrated, the functionalities, or some of the functionalities may be implemented as an integrated functionality.

The network 102 may comprise one or more networks, which may be of same type or different type. Typically social media servers 120 are connected to an Internet Protocol-based network, and they use protocols like http (Hypertext Transfer Protocol) or https (Hypertext Transfer Protocol Secure) for transmission.

The social media server 120 represents one or more network entities providing a social media service. A social media service, also called a social media, a social media network, a social media networking service, may be defined as a connectionless application/platform which a user uses to publish content that is typically, but not necessarily, by default public at least to other users of the specific social media service or "followers", "friends" or "connections" of the user. For the Industrial Internet of Things, the social media service is preferable but not necessarily a private social media service providing private communication within or- ganizations or between organizational members and pre-designated groups. The difference between the public and the private social media service, also called enterprise social media service, is that to obtain access to the private social media service, the entity wanting to obtain access has to be accepted by an authority in the private social media service in advance, for example by giving an email account in the domain used by the private organizations or by sending an invitation to join, whereas in public social media service at most a registration is needed. Another way to define the difference is that a private social media service is a closed service, typically organisation-specific or enterprise-specific, whereas a public social media service is an open service, or a semi-open service, typically a cross-organisational service. Examples of the private social media service include Yammer, TIBCO tibbr, Socialcast, and Skype for Business (earlier Microsoft Lync). WhatsApp, supporting closed user groups, is also a kind of private social media. It should be appreciated that the amount of social media services including web-based public and private services, like the above identified micro-blogging services, and social status update publication applications, is evolving and the above list is not an exhaustive list.

Below different examples are described assuming that a post is a content that is published on an account in the social media server, without restricting the examples to such a solution. The account used may be the user's account, or an account registered to the analyser apparatus, or an account registered to an analytic tool. The analytic tool account may have been registered analytic tool -specif- ically enterprise-specifically (in which case access information is preferably installed with the analytic tool) or analyser apparatus -specifically (in which case access information is preferably stored to the analyser apparatus when the account was created), for example. Other ways to register the analytic tool account may be used as well. It should be appreciated that the content may be textual, visual or aural content or any combination thereof. Further, it is assumed that the social media service supports a possibility to users to track (follow) posts as well as search them with specific keywords. A keyword is formed by tagging, for example with a hash symbol (#), a specific word. Yet another assumption made, for the sake of clarity, is that it is assumed that there are no restrictions to length of posts. If there are, it is obvious for one skilled in the art how to divide a longer post to a series of posts whose length is within the length limits.

Figure 2 shows an exemplary information exchange in a system having devices Dl and Dn in the Industrial Internet of Things, an analyser apparatus AS with external memory MEM that is accessible by both a social media server SOME and the analyser apparatus AS. Further, two users are providing user inputs via user interfaces Ul and U2. A further assumption in Figure 2 is that the same user information is used to obtain access to the analyser apparatus and to the social media service, without restricting the example to such a solution. Therefore in the illustrated example, the access procedures are not described; the social media unit is configured to sign in automatically, without user involvement, to the social media service, using the user's access information. Correspondingly a user signed in the social media service may view, without user involvement, data in the analysis system. It should be appreciated that the social media unit, or access monitoring system/application, may also be configured to prompt the user for access information, at least in cases where the user's account is used.

In Figure 2, Dl sends raw data in message 2-2 to AS. AS determines the type of Dl, and retrieves, by sending message 2-3, from MEM, or more precisely from data model definitions in MEM, those data model definitions that are to be used with the type of Dl.

When AS receives the data model definitions in message 2-3, AS indexes in point 2-4 the data received in message 2-2. The indexing means that data context, as defined by the data model definitions, is associated with the received measurement data, also called a raw data, and further associated with a pointer pointing to a memory storage area whereto at least the measurement data, possibly with the data context, are stored (message 2-5) During the association an entry to the list for the Industrial Internet of Things context and the corresponding measured data is created.

Also Dn sends, in message 2-6, raw data to AS, which then performs the above described functionalities (messages 2-2', 2-3', point 2-4' and 2-5'). It should be appreciated that Dn may be another type of device or process, and hence the data model definitions may be different causing the data context to be stored to be different, also in other respects than mere values.

In the illustrated example, a user then starts an analytic tool by providing via a user interface Ul a corresponding instruction 2-7 or instructions. In response to receiving the instruction, AS processes in point 2-8 the instruction. More precisely, AS detects the analytic tool selected, initiates the analysing by configuring the analytic tool to correspond user settings, for example, retrieves (messages 2-9, 2-10) from MEM one or more data sets needed, according to the configuration, for example, by the analytic tool to perform the instructed data analysis. The analysis itself, including possible storing to the memory, outputs to UI and further in- structions received from the user via UI, are not illustrated in Figure 2.

However, at some point, the user determines that the results of the analysis merit further discussion in and analysis by the social media service. Therefore the user selects a social media service tool, for example by clicking a Yammer tool icon or other selection item outputted on the user interface and being selectable during the analysis. The user input selecting the social media tool is received in instruction 2-11.

In response to receiving the instruction, AS triggers the social media unit that collects in point 2-12 information on analysed data sets, i.e. pointers to the data sets, information on analytic tool currently in use, its configurations such as the analytical method used, further analysis context and information that is to be posted. The information on analytic tool and its configurations is stored as an entry to the list for analysis information. It should be appreciated that in another embodiment, the social media unit may not collect information on the analysis tool configuration in which case the entry does not comprise configuration information. The information on the analytic tool (with or without configuration information) and the data context, i.e. information on the analysed data sets, form an analysis context. Depending on an implementation, the analysis context, collected by the social media unit, may comprise one or more outputs and/or analysis results in addition to the information on the analytic tool and data context or information on data context. Hence, the user does not have to take snapshots or screenshots. Further, the information to be posted is determined, on the basis of the data model definitions and/or data context and/or analysis context and/or possible additional preset settings. The pre-set settings may depend on analytic tool, and/or type of the Things whose data is analysed and/or data model definitions used in indexing. The pre-set settings may define at least part of the information to be posted and/or one or more keywords to be used. The collected information is converted in point 2-12 to a social media service format. More precisely, one or more predefined keywords are created and/or determined from the analysis context and/or data context, and the keywords, possible additional content, a pointer to the used analytic tool and its configuration, and one or more pointers to analysed one or more data sets are determined and combined/added to form the body of the post, i.e. the actual content to be published, and then converted to a post content in the format used by the social media service.

Instead of having all the above information as the body of the post, the above described body of the post, or at least some of it, may be stored to the memory as a posting record, and the body of the post then may contain a link, or a corresponding pointer, to the posting record. Then a header is added and the post is ready.

When the post is ready, the social media unit causes AS to send, or upload, the post in message 2-13 to SOME. Since the post is received as if it would be a human-generated post, no configuration changes are needed for SOME, and therefore the process is not described in detail herein. In response to the post, SOME adds in point 2-14 the content to its database, and makes it available for viewing. Further, SOME acknowledges the post by sending message 2-15, message 2-15 containing a link to the post. If the post contains a keyword someone is fol- lowing or the post is assigned to a particular group, information on the post may be sent to followers, depending on the social media service settings. In response to receiving in message 2-15 the link, the AS associates in point 2-16 the link with the analysis context. This association may be performed by updating the entry on the list for analysis information to include the link. Another alternative is to store the link to the posting record, if such records are created. Yet a further alternative is to associate the link with the pointers to the data sets. By means of this link, any user of the AS, trying to solve the same problem may obtain the information in the social media service. For example, the user may view a discussion. Further, the analytic tool, and/or the social media unit monitoring analytic tools, may be configured to detect a similar situation and obtain, using the link, in- formation from the social media, to include it to the data sets to be analysed by the analytic tool, for example.

At some point, another user notices the post and the problem, wants to study the actual facts behind the problem and selects one or more of the pointers, shown via user interface U2, pointing to the one or more data sets and the analytic tool with the configuration. The data is retrieved (messages 2-17, 2-18) from MEM, either directly, if the post was downloaded to a user apparatus, or via SOME, using the pointers and/or links that were included in the body of message 2-17. Thanks to pointers being collected and published, the other user using social media will be outputted (point 2-19) exactly the same machine-generated information than the user using the analytic tool, and not a user-generated verbal impression of the information and/or static snapshots/screenshots. More precisely, the other user has access to the same data sets, knows the configuration of the analytic tool, and may manipulate the analysis results as if the user were using the analytic tool.

If the other user responds something (message 2-20), it is obtainable in AS via the link stored in point 2- 16 by any user viewing the same analysis. However, that is not illustrated in Figure 2. Naturally, if AS, or the analytic tool has been registered as a follower in the group the post was uploaded, information on responses to the post will be sent to the address used for AS or the analytical tool. The information may be stored to the memory, possibly associated with the original post or its link, for example to expand the knowledge base, as described above, and/or it may generate an alert "further information received", for example.

As is evident, by means of the pointers to the data contexts analytic tool used, and links to the social media (or information copied via the links), two dimensional data searches and analysis may be performed. In other words, data received from the Industrial Internet of Things and data received from users (human individuals) is connected and searchable/analysable in the analysis system side as well as in the social media service side.

It should be appreciated that the raw data indexed may be previously collected raw data, already stored to the memory. For example, Dl could send raw data and indicate the type to be used in the indexing. Hence, the analyser apparatus may use "old data", that has been collected before installing the analyser apparatus, and/or data collected after the analyser apparatus has been installed. Figure 3 illustrates an exemplary functionality of another social media unit. In the illustrated example, the social media unit is running as a background process all the time.

Referring to Figure 3, data analytic tools are monitored in step 301 until a trigger event for the social media service is detected in step 302. A trigger event may be a user input selecting "to social media", like the Yammer button, or an output of an analytic tool may be configured to be a trigger event. For example, a self- learning analytic tool may detect that it cannot identify or correct a root cause, and a corresponding output, like alert via a user interface, is a trigger event. Another example is that a feature vector, calculated from the data in the one or more data sets, is compared to a pre-set target, and if the difference is more than a threshold, a trigger event is detected. The pre-set target may be a combination of measurement values providing a process result meeting quality targets. Still a further example of a trigger event is detecting a scanning observation by an analytic tool scanning enhanced knowledge base, or part of the (enhanced) knowledge base. It should be appreciated that there is no restrictions relating to what constitutes a trigger event. The trigger event may be even a rule, for example the same root cause being detected at least five (or any number) times in a period. Further, an analytic tool may be configured to define and/or update one or more trigger events.

When the trigger event is detected, the analytic tool via which the trigger event was detected, is determined in step 303, as well as the configuration of the analytic tool at the time the trigger event was detected. Then, in the illustrated example, the keywords are determined in step 304, for example as described above, and the pointers to the one or more data sets causing the trigger event, and a pointer to the configuration and the analytic tool are determined in step 305. At this stage at least a preliminary content for the post is ready. In the illustrated example, it is then checked in step 306, whether or not the analytic tool is one of analytic tools for whose posts also user input is asked. That information may be defined in analytic tool configurations, or there may a separate list indicating such analytic tools, for example. Examples of user input that may be asked include con- text tags, i.e. keywords, or information transferrable to one or more keywords, classification of the problem (to which discussion forum the problem should be posted, for example), removal of one or more system generated keywords, acceptance or rejection of the system performed classification of the problem, a commentary on the problem, a root cause hypothesis, and short description of the problem. If the user input is to be asked (step 306), the user is prompted for input in step 307, and after receiving in step 308 the user input for the post, and adapting in step 308 the preliminary content according to the received user input, or if user input is not to be asked (step 306), the social media message (post) is formed in step 309, and it is sent to be published in step 310. Then the process proceeds to step 301 to monitor data analytic tools.

In another exemplary embodiment, no user input is asked, and the steps 306, 307 and 308 are omitted. In a further embodiment, user input is always asked, and step 306 is omitted.

In addition to the above, the social media unit, for example, may be configured to associate the keywords and/or the classification inputted by the user with the one or more data sets analysed and/or with the data model definitions so that in future they will become candidates for keywords suggested by the social media unit.

Figure 4 illustrates another exemplary functionality of a social media unit. The functionality illustrated herein is an additional functionality providing some examples illustrating, how the fact that social media service may have been used, may be utilized in the analysis system side. In the illustrated example, the social media unit is running as a background process all the time.

Referring to Figure 4, data analytic tools are monitored in step 401 until it is detected in step 402 that an analytic tool retrieves data from the memory to be processed (analysed). In response to the data being retrieved, it is checked in step 403, whether or not the retrieved data is associated with one or more links to one or more social media posts. This checking may be performed via an analysis context referring to the same data, via posting records, or via any type of association used for associating a link to the information indicated in the post and maintained in the analysis system.

In the illustrated example, if the retrieved data is associated at least with one link to a post in the social media service, the process takes care in step 404 that a flag, or a tooltip or any corresponding hint, is outputted via a user interface when the data, or analysis result using the data, is outputted so that the flag indicates to the user that for this data set there is at least a post, and possibly further information or discussion, in the social media service. Further, the flag may contain the link to enable easy access to the post, and the discussion.

After that, or if the retrieved data is not associated with a link to the social media service (step 403), it is checked in step 404, whether or not there are similar data sets, used by the same analytic tool, that are associated with one or more links to the social media service. There are no restrictions which data sets may be considered to be similar ones. For example, feature vectors of the data sets may be calculated and compared, and if the differences are between a preset mar- ginal, the data sets are determined to be similar ones. Another example includes use of fingerprints in a similar manner. Further alternative include that the data sets have the same or overlapping parameters, and/or the same type of alert created by the system.

If there are one or more data sets that are similar to the retrieved data set (step 405), at least the links, possibly some additional information, like a question posed, if in the record, for example, is outputted in step 406 via the user interface to the user of the analysis system. By means of obtaining these links, the user receives easily information, whether or not, while trying to solve the problem, one should study also information in the social media service, and if the user decides to do so, the outputted links provide an easy and accurate way to obtain the additional information.

After outputting the one or more links, or if there are no similar data sets (step 405), the process returns to monitoring (step 401).

Figures 5A to 5F shows a use example how data collected and analysed in the Industrial Internet of Things side 501 by one or more devices may be combined with information obtainable via human social media services 520.

Figure 5A illustrates a starting point in which in the Internet of Things side 501, comprises in a memory 511 data 511-1 generated by the Things in the Internet of Things, different analytic tools 511-2, 511-2' (only two are represented) that may be used to analyse the data in the memory, a social media tool 511-3 to support the interaction with the human social media service side 520, and an operator's analysis view 512. It should be appreciated that the amount of data 511-1 may be huge and comprise plurality of different data items, and that the different data items illustrated in Figures 5A to 5G may be only part of the data 511-1 gen- erated by the Things. The operator's analysis view represents herein what is outputted in a user interface by an analysis tool to a user (operator) of the analysis tool. Since in the illustrated example of Figure 5A the operator is not analysing anything the view 512 is empty.

In the starting point of Figure 5A, the human social media service side 520 comprises an empty post area 521 and an empty user's view 522 for a specific post. It should be appreciated that other kind of view/output arrangement may be used as well.

Then a trigger event, for example a device "dev A" has stopped and restarted, has happened. Referring to Figure 5B, the operator has selected analytic tool two and uses it with settings "set one", illustrated by 511-2'a in Figure 5B. The analytic tool outputs on the operator's view 512 results obtained from a data set 511-3, i.e. from the data stored therein. It should be appreciated that the amount of data in the data set 511-3 may be huge and comprise plurality of different data items, and that the data set may be only part of the data 511-1 or comprise the whole data 511-1. The operator spots the peaks in the middle in the operator's view 512, concludes that the peaks are the reason that caused the device "dev A" to stop and restart but the operator cannot determine what causes the peaks. Therefore the operator selects the social media tool 511-3 (the selection is not illustrated in Figure 5B).

The social media tool then determines a pointer that points to the com- bination illustrated by 511-2'a, i.e. pointers to the analytic tool two 511-2' and the set one used, and pointer pointing the data set 511-3 analysed, and determines in the example stop and dev A to be the keywords. Further in the example the social media tool prompts the operator to provide further information, and the operator inputs a question "What causes peaks to dev A?"

Once the information is ready, the social media tool converts the information to a post and posts it in the social media service side 520, the post being shown in the social media area 521 as a post 521-1 in Figure 5C. In other words, the machine-to-machine device comprising the social media tool in the Industrial Internet of Things side acts as if it were a combination of a user forming a very specific post to a social media service and a user equipment of the user posting the post. Further, the social media tool adds to the data 511-1 in the memory 511 a link to the post 521-1. In the moment illustrated in Figure 5C the operator is not any more analysing, and there are no user in the social media side 520, and therefore views 512, 522 are empty in Figure 5C.

Then the post in the social media has been noticed by a user. The user has clicked the combination formed by the pointers in the post 521-1. As can be seen from Figure 5D, clicking the pointers causes the social media service side 520 to provide the user exactly the same output on the user's view 522 that was out- putted (see Figure 5B, view 512) to the operator, by retrieving (502') the data and the analytic tool or its information with its setting (502) or at least the setting, as is described above. Further, the pointers provide the user a further possibility to study the data set, and if the implementation allows, also around the data set. In prior art solutions that is not possible. The view could be the same only if the operator had taken a snapshot on the view and posted it, but then there would have not been a possibility to manipulate the data.

In the illustrated example, the user realizes that the problem is the same what they experienced a week ago: there were some dust in a rear fan and once the dust was removed, the problem was solved. Therefore the user replies in the social media area to the post, the short reply 521-2 being illustrated in Figure 5E in a situation in which neither the operator nor the user are analysing. It should be ap- predated that the user may have been aware of a similar problem discussed elsewhere, so he/she may have added to the response a link to such a discussion. For example, the response might have been "there may be dust in a rear fan, but if removing the dust does not solve the problem, further solutions can be found in discussion A" (and A has the "link" to the discussion).

The operator notices the post, dusts the rear fan, and then notices that the problem is solved. In the illustrated example, the operator uses his/her social media service and posts to the social media service a reply 521-3 indicating that the problem was solved, adding an hashtag "db". In the illustrated example, it is assumed that the social media tool 511-3 is configured to add to the data 511-1, preferably associate with the link, a summary of the discussion in the social media in response to detecting in the social media service the hashtag "db", as illustrated in Figure 5F, so that the data comprises in addition to the data generated in the Internet of Things side data created in the social media side. In the illustrated example, the data created in the social media side includes nickname of the user (@master), the data providing root cause (dust in rear fan), the problematic situation defined by the data set 513, and the analytic tool used with its settings used, and then the keywords. Naturally, there is no obligation to the operator to post the result 521-3, and/or the summary to be added to the data in the Industrial Internet of Things side.

As can be seen, by combining machine-to-machine networking and human-to-human networking the enhanced knowledge database may comprise both measurable information and information that cannot be measured, such as silent human knowledge.

Figure 5G illustrates an enhanced knowledge database 530. As is described with Figure 5D, the data 521 in the social media service side is enhanced, thanks to the pointers, to cover data in the Industrial Internet of Things side 501. As is illustrated in Figure 5D, the data in the Industrial Internet of Things side 501 is also enhanced (502") to cover discussions created by posts initiated by the social media tool in the Industrial Internet of Things side 501. To utilize the enhanced knowledge database different kind of enhanced search tools 511-4 (or correspond- ing units) may be created to be used by analysing apparatuses 510, or corresponding devices. In an enhanced search tool, the search may be based on machine learning, results may be outputted even when the match is not a perfect match but, when combining the results in the enhanced knowledge database, an adequate enough match is noticed, the result may be outputted. Further, the enhanced search tool may be learned by collecting use information, for example by storing search information (search criteria, search results) and related user's selections to the enhanced knowledge database.

One use example is that an operator is analysing, by an analytic tool, a data set amongst the data collected from the Industrial Internet of Things, and wants to find the best experts in the field relating to the context of the analysed data set. The operator selects the enhanced search tool, provides as a search criterion "Find a best expert", and that triggers the social media tool to create pointers and keywords, which are posted to the social media service. In addition, the enhanced search tool uses the keywords and/or the pointer information as additional criteria, and searches the data 511-1 and the linked social media service discussions in the social media service to find the best expert. For example, the @master may be identified by the enhanced search tool to be the best expert since his answer solved the problem.

The enhanced search tool may be configured to analyse and/or visualise posts created by social media tools in the Industrial Internet of Things side to find certain patterns, as is done by different analysing and/or visualisation tools in the social media service side, for example to identify trends, identify device types or environments causing problems, identify persistent problems, output a time dependent map of posts showing the evolution of a problem and thus helping with problem diagnosis / root cause analysis, to identify compromised nodes in the event of a cyber-security incident, etc. Naturally, thanks to the enhanced knowledge database and the pointers, the different analysing and/or visualisation tools, like topic modelling, network mapping, in the social media service side will have information also from the Industrial Internet of Things side, which in turn increases the quality and/or predict- ability of the analysis.

As is evident from the above, the social service unit provides a tool that, depending on a solution, replaces or assists human activities by determining pointers and at least part of the keywords without user involvement and without affecting to the actual functionality of the analytic tool used or of the social media service used. In other words, it provides a systematic mechanism to combine machine-to- machine networking and human-to-human networking, which may be called an Internet of Things, services and people. This results in faster resolution of the issues, especially new unforeseen issues, enhanced quality of solutions found, and reduced used of human labour for inputs, all minimizing the time the monitored pro- cess/system does not function properly, thereby increasing overall productivity and overall equipment effectiveness. Further advantages include better usability of a production (manufacturing) process, increased safety, and more properly targeted maintenance.

The steps, points, messages (i.e. information exchange) and related functions described above in Figures 2 to 5G are in no absolute chronological order, and some of the steps/points may be performed simultaneously or in an order differing from the given one. Other functions can also be executed between the steps/points or within the steps/points, and other information may be sent. For example, access information may be requested. Some of the steps/points or part of the steps/point can also be left out or replaced by a corresponding step/point or part of the step/point.

Further, it should be appreciated that a social media analysis tool, analysing data in the social media, may be configured to utilize information stored to the memory in the analysis system. For example, the social media analysis tool may search for a malfunction that has raised lot of concern among users, and then looks from the analysis system a root cause, or root causes. The links in the social media posts enable this.

The techniques described herein may be implemented by various means so that an apparatus implementing one or more functions described with an embodiment, or a combination of embodiments, comprises not only prior art means, but also specific means for implementing the one or more functions described with an embodiment and it may comprise separate means for each separate function, or specific means may be configured to perform two or more functions. The specific means may be software and/or software-hardware and/or hard- ware and/or firmware components (recorded indelibly on a medium such as readonly-memory or embodied in hard-wired computer circuitry) or combinations thereof. Software codes may be stored in any suitable, processor/computer-read- able data storage medium(s) or memory unit(s) or article(s) of manufacture and executed by one or more processors/computers, hardware (one or more apparat- uses), firmware (one or more apparatuses), software (one or more modules), or combinations thereof. For a firmware or software, implementation can be through modules (e.g., procedures, functions, and so on) that perform the functions described herein.

Figure 6 is a simplified block diagram illustrating some units for an ap- paratus 600 configured to be an analyser apparatus, or a corresponding apparatus, comprising at least the social media unit or corresponding functionality or some of the corresponding functionality if functionalities are distributed in the future. In the illustrated example, the apparatus comprises one or more interfaces (IF) 601 for receiving and/or retrieving and/or transmitting information both to the social media service and to the Industrial Internet of Things, and to/from an external memory, if such is used, and/or outputting and receiving information from a user, a processor 602 configured to implement at least the social media unit, described herein, or at least part of corresponding functionality as a sub-unit functionality if distributed scenario is implemented, with corresponding algorithms 603, and memory 604 usable for storing a computer program code required for the social media unit, or a corresponding unit or sub-unit, i.e. the algorithms for implementing the functionality. The memory 604 is also usable for storing other possible information, like the lists, raw data, records, etc.

In other words, an apparatus configured to provide the analyser appa- ratus, or an apparatus configured to provide one or more corresponding functionalities, is a computing device that may be any apparatus or device or equipment or node configured to perform one or more of corresponding apparatus functionalities described with an embodiment/example/implementation, and it may be configured to perform functionalities from different embodiments/examples/imple- mentations. The social media unit, as well as corresponding units and sub-unit and other units, like the data storing unit, described above with an apparatus may be separate units, even located in another physical apparatus, the distributed physical apparatuses forming one logical apparatus providing the functionality, or integrated to another unit in the same apparatus.

The apparatus configured to provide the analyser apparatus, or an ap- paratus configured to provide one or more corresponding functionalities may generally include a processor, controller, control unit, micro-controller, or the like connected to a memory and to various interfaces of the apparatus. Generally the processor is a central processing unit, but the processor may be an additional operation processor. Each or some or one of the units/sub-units and/or algorithms de- scribed herein may be configured as a computer or a processor, or a microprocessor, such as a single-chip computer element, or as a chipset, including at least a memory for providing storage area used for arithmetic operation and an operation processor for executing the arithmetic operation. Each or some or one of the units/sub-units and/or algorithms described above may comprise one or more computer processors, application-specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field-programmable gate arrays (FPGA), and/or other hardware components that have been programmed and/or will be programmed by downloading computer program code (one or more algorithms) in such a way to carry out one or more functions of one or more embodiments/implementations/exam- ples. An embodiment provides a computer program embodied on any client-readable distribution/data storage medium or memory unit(s) or article (s) of manufacture, comprising program instructions executable by one or more processors/computers, which instructions, when loaded into an apparatus, constitute the social media unit, or its sub-unit. Programs, also called program products, including software routines, program snippets constituting "program libraries", applets and macros, can be stored in any medium and may be downloaded into an apparatus. In other words, each or some or one of the units/sub-units and/or the algorithms described above may be an element that comprises one or more arithmetic logic units, a number of special registers and control circuits.

Further, the apparatus configured to provide the analyser apparatus, or an apparatus configured to provide one or more corresponding functionalities may generally include volatile and/or non-volatile memory, for example EEPROM, ROM, PROM, RAM, DRAM, SRAM, double floating-gate field effect transistor, firmware, programmable logic, etc. and typically store content, data, or the like. The memory or memories may be of any type (different from each other), have any possible storage structure and, if required, being managed by any database management system. In other words, the memory, or part of it, may be any computer-usable non- transitory medium within the processor/apparatus or external to the proces- sor/apparatus, in which case it can be communicatively coupled to the processor/apparatus via various means as is known in the art. Examples of an external memory include a removable memory detachably connected to the apparatus, a distributed database and a cloud server. The memory may also store computer program code such as software applications (for example, for one or more of the units/sub-units/algorithms) or operating systems, information, data, content, or the like for the processor to perform steps associated with operation of the apparatus in accordance with examples/embodiments.

It will be obvious to a person skilled in the art that, as technology ad- vances, the inventive concept can be implemented in various ways. The invention and its embodiments are not limited to the examples described above but may vary within the scope of the claims.

Claims

1. A computerized method comprising:

running in an apparatus an analytic tool analysing data that comprises data generated by Industrial Internet of Things;

detecting a trigger event for a social media service dedicated to human interactions;

performing, in response to the trigger event, by the apparatus, at least the following:

determining one or more first pointers pointing to one or more ana- lysed data sets analysed by the analytic tool and a second pointer pointing to analysis information indicating at least the analytic tool;

determining one or more keywords;

converting the one or more first pointers, the second pointer and the one or more keywords to a post in a format used in the social media service; and sending the post to a social media server providing the social media service.

2. A computerized method as claimed in claim 1, further comprising: determining analytic tool configuration used by the analytic tool when the trigger event was detected; and

adding information on the analytic tool configuration to the analysis information.

3. A computerized method as claimed in claim 1 or 2, further comprising:

determining one or more of the one or more keywords based on the contexts of the one or more analysed data sets.

4. A computerized method as claimed in claim 1, 2 or 3, further comprising:

prompting via a user interface a user to provide as a user input at least one of further information and one or more keywords; and

adding received user input to the post.

5. A computerized method as claimed in claim 1, 2, 3 or 4, wherein the trigger event is a user input indicating a social media service.

6. A computerized method as claimed in claim 1, 2, 3 or 4, wherein the trigger event is an output of the analytic tool.

7. A computerized method as claimed in any preceding claim, further comprising: receiving a link to the post published in the social media service;

storing the link; and

associating the link with at least one of the analysis information in the post, the one or more analysed data sets and a post record.

8. A computerized method as claimed in claim 7, further comprising: outputting the link in response to detecting the same or similar analysis situation.

9. A computerized method as claimed in any preceding claim, wherein the data further comprises at least one of links to social media service discussions and copies of the discussions, the data thereby forming an enhanced knowledge base.

10. A computer program product comprising computer program code configured to perform a method as claimed in any one of claims 1 to 9 when executed on a computing device.

11. An apparatus comprising means for performing a method as claimed in any one of the claims 1 to 9.

12. An apparatus as claimed in claim 11, the apparatus comprising at least one processor and at least one memory including a computer program code, the at least one memory and the computer program code configured to, with the at least one processor, to provide the means for performing a method as claimed in any one of the claims 1 to 9.

13. A system comprising:

Industrial Internet of Things comprising sensors configured to transmit data;

one or more memories for storing at least data received from the Industrial Internet of Things;

one or more servers providing a social media service dedicated to human interactions; and

one or more apparatuses as claimed in claim 11 or 12.

14. A system as claimed in claim 13, the system being configured to utilize both the data in the one or more memories and data associated in the one or more servers with posts sent by the one or more apparatuses and discussions relating to the posts as an enhanced knowledge database.