Content for  TR 22.856  Word version:  19.2.0

Artificial Intelligence (AI) is becoming more and more popular in many areas where especially humans cannot handle complicated tasks well (e.g., factory, vehicle, robot, mobile). This trend is likely to continue, and AI will be applied to even more areas. In addition to the rapid expansion of AI, the AI technology itself is also improving. AI that can express emotions like humans and AI that can communicate naturally are now emerging. Given these trends, AI could one day be used not only for industrial use cases, but also as our personal partner and personal assistant to perform many of the tasks around us. This use case proposes a communication with an autonomous virtual alter ego, which is an AI-based digital representation acting autonomously on behalf of a user herself/himself in the mobile metaverse services. For example, user's autonomous virtual alter ego autonomously sends a mail to clients on user's behalf. Also, the alter ego can autonomously communicate with the user, other physical users, and other alter ego by using the network capabilities based on the user's 3GPP subscription. Therefore, the use of network by the alter ego has to be captured correctly by the network from charging point of view. All the experience and knowledge performed both in physical world and metaverse will be shared between the alter ego and its user, thus creating more than double the opportunities to play multiple roles simultaneously. This autonomous virtual alter ego concept aims to improve emotional well-being, health, and life satisfaction by enabling users to perceive many opportunities in life, such as balancing work and family and participating in many communities simultaneously.

John has a UE which has connectivity to 5GS based on subscription to MNO and a contract with an autonomous virtual alter ego service provider. The service setting and parameters for this alter ego service are stored in the user's subscription data. John's virtual alter ego has been trained by the autonomous virtual alter ego service provider, so he can enjoy the alter ego application via his UE. There are two kinds of application servers. One is for alter ego application. The other is for other applications which the alter ego application connects for executing tasks. The autonomous virtual alter ego service provider is trusted by MNO, and the alter ego can use the network capabilities autonomously based on user's subscription to MNO. The MNO offers a service enabler that allows John to request that the network limit how much resources his alter ego is able to consume on his behalf. The service enabler also provides John with storage space that he can use to store application specific data and information about himself in the network. John is able to configure the enabler to give the virtual alter ego limited access to John's information.

1. John has a F2F appointment with his important client at client's office, so he cannot attend an internal web meeting scheduled on the same time. Then, he connects to the alter ego application via his UE and asks his alter ego to show up and participate in the internal web meeting on his behalf.
2. His virtual alter ego checks the network resources on 5GS and computing resources on the alter ego server. Then it judges whether the requested task is able to be process or not. The virtual alter ego checks with the enabler server to see what information about John it is able to access.
3. If the alter ego can complete the tasks, the information from the enabler server is used to train the alter ego and the alter ego starts the tasks. Otherwise, the alter ego proposes task examples it can do with current resources to him, so that John can reconsider the request. For example, the alter ego can only listen during the meeting and take some notes but can't say anything. Once John and the alter ego agree on what task(s) the alter ego will perform, John silences, or turns his off his UE so that he can focus on the F2F meeting.
4. Before the alter ego attends the meeting, it accesses the web meeting server as John's alter ego via internet or 5GS by receiving the permission from John and web meeting server. When the company internal information or some other information is needed for the meeting, the alter ego asks the enabler to permit access to the information or asks John to permit the access to them.
5. At the meeting, the alter ego explains something, makes some questions/comments to other attendees (including physical humans and other virtual alter egos).
6. After the meeting, the alter ego autonomously reports to John via 5GS by message or on call. He looks or listens to the report and returns feedback and new requests by message or on call if any.
7. When the alter ego communicates via 5GS, the charging information is collected and linked to user's charging information to charge the user subscribing the alter ego service.

After receiving the feedback from John, the alter ego is retrained. Then, the autonomous virtual alter ego becomes more accurate and finishes the tasks much more immediately. As a result, the time available to him in life is more than doubled.

AIML model transfer frameworks documented in TR 23.700-80 can be applied to this use case. IMS MMTEL services documented in TS 22.173 can be applied to this use case.

[PR 5.17.6-1] [PR 5.17.6-2]