Elderly populations are increasing in Norway, Scandinavia and other developed countries, in part because people are living longer due to medical advances. This is associated with several societal challenges, including an increasing demand for nursing homes, which may soon outstrip supply, and a projected nurse shortage. Moreover, some older people prefer to ‘age in place’ – to stay in their own homes in safe and dignified living conditions for as long as they can take care of themselves – and welfare technology may help make this more possible. The specific type of welfare technology researched in this thesis is in the area of human behaviour modelling (HBM) and represents a relatively new area of research. HBM seeks to model the behaviour of a person living alone in a smart environment in order to detect abnormal behaviour and alert family members or caretakers if something is wrong. It is based on an assumption that people tend to follow specific behavioural patterns in their daily lives. HBM should be tailored for each individual user since people have unique behaviour patterns. In the present thesis, a behaviour is defined as a combination of activity, posture, location and duration. Abnormal behaviours include, but are not limited to, falls and early signs of cognitive impairment. This thesis analysed several algorithms to detect abnormal behaviour: decision trees, the hidden Markov model (HMM) and the hidden semi-Markov model. HBM has been developed and tested using a real-world, open-source dataset. The successful application to welfare technology requires consideration of a number of additional ethical and legal aspects. In addition, older people’s attitudes towards welfare technology must be taken into account during the research and development phases to reduce the risk of rejection from its intended end-users and to ensure a person-centred approach to integrating new technology. This thesis therefore consists of two parts: a main technical part, which discusses the technological development of HBM, and a health care part, which discusses HBM’s ethical and legal implications as well as older people’s opinions about the use of HBM in welfare technology. This thesis includes four Journal Articles and four Conference Articles. Overall, their results showed that it was possible to model an individual’s behaviour and detect abnormalities using statistical models. The best results were obtained using HMM, which successfully detected abnormal behaviour such as falls, and changes in the duration of behaviours performed by an individual. In addition, the research examined opinions about the use of HBM for welfare technology among older people living in Norway. Most participants expressed that they wished to maintain their independence and autonomy, to feel safe in their own homes and to age in place, and they expressed positive opinions about the use of HBM and the great convenience it offered. Surprisingly, they expressed no concerns about privacy. Although a few mentioned concerns about loss of autonomy and dignity, most participants indicated that the potential benefits of HBM outweighed their concerns.