Discovering Potential Communities of Practice with Social Intention in an IoT-based Smart Home Environment

Dongman Lee
November 11, 2014


With the proliferation of IoT, multiple smart homes equipped with IoTs can interconnect with each other and form a community which we will call a smart home community. With the aid of a smart home community, we can discover potential social proximity among users and generate communities of practice that increase the quality of life for community members. In this article, we introduce the current state-of-the-art and identify some remaining research issues.

The proliferation of IT devices allows users to achieve far more complicated tasks than ever before by following the deep knowledge work approach introduced by P. F. Drucker in 2003. Users can come together to learn and share information and form a "Community of Practice" for improving their productivity. In particular, there is an emerging need for communities of practice to improve the quality of life by communal sharing of tasks such as child care and cleaning the environment of apartments as a large portion of urban dwellers live in high-rise downtown housing. Helping users to form a variety of communities of practice for enhancing the quality of life is one of the important challenges in smart home research.

Traditional smart home solutions help users to find various service opportunities based on occupant intelligence by providing flexible and adaptive task environments and control capability. Furthermore, the proliferation of IoT makes it possible to interconnect multiple smart homes over the internet. In such environments, it is possible to provide services for multiple families by forming a community of practice according to their intentions, which we call Smart Home Community. For example, Smart Community Architecture connects multiple smart homes by interconnecting home gateways for providing various community services such as neighborhood watch, pervasive healthcare, etc. (Li et al., IEEE Communications Magazine, November 2011). According to this trend, a new challenge is how to discover and form appropriate communities of practice for users and provide services for them using, especially, IoTs available nearby.

In traditional living environments, people can get such community support by forming an intentional community which is a group of people who have chosen to live together with a common purpose, working cooperatively to create a lifestyle that reflects their shared core values. For example, cohousing, which is an example of an intentional community, delivers economic benefit to the users by sharing living environments. Although the advantages of intentional communities are clear, in practice it can be difficult to populate intentional communities. For example, community members should live close to each other to make collaboration easy and that makes it hard to form a different community for temporal user needs. However, with the aid of smart home communities emerging from the population of IoTs, we no longer need to be concerned about physical proximity constraints. We could form a variety of smart home communities whenever any particular smart home community is required. For example, we could reduce the baby care overhead by activating social collaborations among neighbors who have common baby care problems. Smart home community-based services could allow collective child care among neighbors. They could then resolve unexpected problems such as helping neighbors to collect their children when they have an emergency task at work or giving appropriate advice on the fever of neighbor's children.

The biggest difference between the traditional smart home and the smart home community environment is that the target of a service is a community whose members stay in physically separated environments. Thus, we need to discover suitable community members who share common interests in baby care at separate smart homes and provide knowledge sharing, emergency, and baby care labor-sharing services for babies of the discovered community members using IoTs. For discovering the baby care smart home community, we need to infer users' intentions on baby care from real-time IoT data (i.e., a wearable sensor data for a baby, an indoor camera, and cooking info from a microwave), babies' life logs from multiple smart homes, and personal web data (e.g., Facebook, Twitter, etc.); discover potential communities for users based on their intentions and the situation of their babies; and ask members of each potential community to take care of their children. For example, Sam has emergency tasks from his boss and cannot collect his son from school so he asks appropriate neighbor Emma to help. It is impractical to assume every community member knows how to contact Emma and direct her to the child. Therefore, the smart home community should be supported by video conferencing services for connecting all community members to let Emma find the child and navigate to Sam's home. For child care (i.e., feeding) Emma should be allowed to access some of the appliances in Sam's home and use those appliances without prior experience of them. Since it is impractical to assume that every smart room has the same set of services with the same interface, the smart home community should support Emma by identifying communication problems and solving them with an appropriate adaptation.

In addition, we need to consider users' willingness to collaborate with each other, privacy, and trust issues for realizing smart home communities. Since a smart home community is activated by user's acceptance, no community can be activated when there is no willingness of users to collaborate. For example, if the smart home community is built upon a give-and-take relationship, the giver may not be interested in the activation of the community. Thus, we need to consider appropriate incentive mechanism(s) to encourage users' participation in the smart home community. The privacy issue is also a critical obstacle to the formation of a smart home community. Since collecting user experiences is important to infer user intentions and form an appropriate community, it is necessary to anonymize the user experiences before collecting them without losing important information. Also, since the community service encourages users to interact with each other, the trustworthiness of each member in the community should be guaranteed to ensure the safety of interactions.



Dongman LeeDongman Lee is a professor in the Department of Computing Science at Korea Advanced Institute of Science and Technology (KAIST) and the Dean of the Graduate School of Culture Technology. Previously, he worked on ubiquitous computing middleware for service continuity across multiple smart home and urban environments. In 2011, he started research on IoT-based cyber-physical service composition and adaptation and elaborated the idea for IoT-based smart home communities.