Introduction
           I will briefly cover the following themes in relation to lessons for IoT standards
          Simplifying applications that               Scaling up for interoperability 
            work across heterogeneous IoT                across multiple communities
            ecosystems, standards and                  Ecosystems of value chains 
            technologies                                 rather than data spaces
          When Digital Twins are, and
                                                       Mimicking how we perceive 
            are not, the solution                        and interact with the world
          Why it is important to keep data 
                                                       Europe leading the world in AI
            and metadata together
                                                                                                2
              Simplifying applications that work 
                            across heterogeneous IoT 
                                 ecosystems, standards
                                          and technologies
            There are many different technologies and                    The Web of Things associates Digital Twins with URIs 
              standards for the IoT                                        (web addresses) and rich descriptions (in JSON-LD) 
            This creates additional challenges for                        covering:
                                                                             The affordances exposed by digital twins in terms of object 
              developers who need to create services that                    properties, actions and events, along with their data types
              involve a multiplicity of IoT ecosystems,                      The semantics and related metadata, e.g., the kinds of 
              technologies and standards                                     things, their relationship to other things, and privacy 
            W3C’s Web of Things has been designed to                        metadata
              simplify these challenges, enabling developers                 Communications and security metadata which is used by 
              to interact with Digital Twins independently of                the client platform to determine how to communicate 
              the underlying IoT ecosystems, standards and                   securely with IoT servers on behalf of the client 
              technologies                                                   application
                                                                          For more details, see:    https://www.w3.org/WoT/
            The Web of Things is ubiquitously applicable 
              anywhere from the network edge to the cloud
                                                                                                                                 3
                                  When Digital Twins are,
                               and are not, the solution
               There is a great deal of hype around the idea                            Live access to sensors and actuators isn’t always 
                 of Digital Twins                                                           what’s needed
               These are software objects that stand in for                                  The GATEKEEPER project centres on improving the 
                 physical objects                                                              care of elderly and frail patients at home. The project 
                                                                                              was predicated on the use of the Web of Things.
                    Digital Twins with live connections                                       However, doctors are more interested in how the 
                       • Applications interact with local software objects, whose              patient has fared over time rather than an IoT centric 
                         implementation hides the underlying details of standards,             perspective of live access to digital twins, which is 
                         protocols and technologies needed to access the physical 
                         objects                                                               better suited to intensive care.
                   Digital Twins as record keepers                                           This implies the need to focus on ingesting and 
                       • Access to historical record of actual behaviour and service           processing heterogeneous information, keeping data 
                         history (object memories)                                             with metadata, and support for proactive and reactive 
                   Digital Twins as simulators                                                processes.
                                                                                              This is when the flexibility of graph databases provides 
                       • Used for testing against simulations of behaviour, e.g. for 
                         developing plans                                                      a better solution
                                 Need for greater focus on graph databases and consideration of new technologies
                                                                                                                                                             4
         Why it is important to keep data
                     and metadata together
         When processing data it is                   We therefore need standards for 
           important to keep  the connection            a wide variety of metadata, 
           between data and its metadata                including contractual and 
         This is critical to preserving privacy        regulatory requirements
           and for respecting terms and                Metadata needs to be 
           conditions of use                            considered when transforming 
         Different sensors may vary in their           data, e.g. whilst anonymising 
           accuracy, and some may be faulty             personal information when 
         You need to know what a given                 sharing data with others – how 
           sensor is measuring (e.g. the                should metadata be transformed 
           temperature of what?)                        in such cases?
                                                                                               5
              Scaling up for interoperability
                 across multiple communities
           Another theme is scaling across multiple              When translating from a rich vocabulary and 
             communities with different use cases, different        data set into one that is weaker, you may be 
             requirements, different traditions and different       fine, but for the other way around, you may 
             terminologies                                          find that you are missing required data and 
           Expecting everyone to use the same ontologies is        metadata
             like trying to force square pegs into round holes    You could supply default values when available, 
           This matters when you want to combine data              but you should enrich the metadata to indicate 
             from heterogeneous sources                             what was missing and what you did about it
           Simple cases include transforming between             Use of intermediate ontologies, analogous to 
             different units of measure                             translating from Burmese to Gujarati via 
           More generally, data mappings will be context           English
             dependent, providing yet another reason to keep      This calls for new standards for mappings 
             the connection between data and metadata               rather than lazily assuming everyone will use 
                                                                    the same ontologies
                                                                                                                    6