From product to service – strategies for upscaling smart home performance monitoring

ABSTRACT Smart home technologies provide an opportunity to address the housing design or retrofit-performance gap and to improve the building regulations. It is currently used to manage comfort, security, and resource efficiency but, adoption remains piecemeal, and disparate. This study aims to explore how householder perceptions of housing quality, and the cost–benefit of improvements informs the adoption of smart technologies. Further, it bridges the theory-to-practice gap by proposing the product-as-service domains that can be deployed for the upscaled implementation of smart home performance monitoring. The survey method returned 972 nationally representative responses. Factor analysis was then used to establish the housing quality priorities, and home improvement drivers that combine to inform the adoption of smart home performance monitoring. Findings show that householders will adopt technologies in return for a ‘benefit’ if reliable smart systems and data-feedback mechanisms are packaged as: (a) commerce services to support utility efficiencies and cost savings, (b) convenience and control services to improve comfort and wellbeing, (c) information and communication services to inform behaviours and decisions, and (d) entertainment services to satisfy hedonic needs. The paper concludes with practical, scalable implementation strategies for smart home performance monitoring. Thus bridging the gap between theory and practice.


Introduction
The UK's Committee for Climate Change (CCC, 2019) recently declared that UK homes are not fit for the future because carbon reductions have stalled, and efforts to adapt the housing stock are not keeping pace with the increasing risks from the changing climate; at the detriment of health, wellbeing, and comfort, including of vulnerable groups like the elderly and those living with chronic illnesses.The drive towards Net Zero Homes (NZHs) represent an effort to address these issues amid rising prices, energy supply volatility, and skills shortages.Homes contribute 22% of UK carbon emissions, but gaps persist between simulated and actual building, occupancy and use performance (Cholewa et al., 2020;Miller et al., 2021).As such homes could use up to 250% more energy than predicted by computer models at design stage (Mitchell & Natarajan, 2020) and conflicting standards mean that a 2020 Part L-compliant house could potentially be less energy efficient than one built in 2013 (Leti, n.d.).Up to 50% difference in energy consumption between the prediction performance and the reality in a retrofit has also been found with very little knowledge on the reasons why (Sunikka-Blank & Galvin, 2012).Emerging evidence also suggest that performance gaps increases once buildings are occupied (Kjaergaard et al., 2020;Rogage et al., 2020).The current drive towards net zero homes necessitates the need to address these issues by investigating social and technical factors together i.e. the indicators that influence prevalent and predominant behaviours during occupancy and use of systems and technologies in the home (Boardman et al., 2005;Chadwick et al., 2022;Schot et al., 2016;Sovacool & Furszyfer Del Rio, 2020).
Housing in the UK is typically delivered through three main providers: private enterprise that are profit driven, housing associations, and local authorities that operate profit and non-profit models.Together, housing associations and local authorities build around 20% of all new dwellings in the UK, circa 42,100 homes in 2019 (Maslova & Burgess, 2023).However, the housing market framework demands little commitment to lifecycle performance beyond the warranty / liability period.Whereas other provisions that facilitate the usefulness and enjoyment of the home such as energy and water, telephone, entertainment -TV and streaming services, and internet services are typically provided by private companies using a service-based approach.The failure in improving building performance, through widescale monitoring has been attributed to the lack of government mandate in initiatives, standards, and policies; building contractors, local authorities, and housing providers with different performance priorities, operating independently of each other; the education sector not promoting building performance in teaching and research; and design teams not working collaboratively with the client/end-user (Stevenson, 2019).Never has there been a greater mismatch between technocratic policy approaches and user service-oriented needs and behaviours (Chadwick et al., 2022).Instead the energy, housing, and utility sectors, which includes governments, regulators, and private organizations, all work towards similar goals but from a myriad of directions.Leading to the design and implementation of policies, engagement, and efficiency strategies that confuse or overlook the full range of influences on household agency, decisions, and actions.Thus, Boardman et al. (2005) argued that improving housing quality requires a multifaceted approach that includes social and market transformations.Social transformation entails understanding the role of the households in securing emissions reductions.It starts with creating opportunities to understand occupants and household behaviours and characteristics as these drive domestic energy and water consumption (Liao & Chang, 2002).Market transformation refers to innovative and transformative policies.It is also key to improving the corporate, social, and environmental image of a sector that is commonly perceived as delivering houses as quickly and cheaply as possible, making profit whilst fulfilling the minimum thresholds of regulatory standards, if at all (Gallent et al. 2010;Maslova & Burgess, 2023).
Smart home technologies proffer opportunities to address the mismatch between housing quality and performance by aligning it with other service provisions in the sector.This study hypothesizes that this alignment can be achieved through (a) the linkage of smart technologies with effective performance monitoring and (b) the transition from a product-oriented housing model to service-oriented approach i.e. the Productas-a-service (PaaS) approach.PaaS was introduced by Goedkoop et al. (1999, p. 3) as a marketable set of products and services capable of jointly fulfilling a user's need.The idea is that value is not necessarily embedded in physical products but in the use of those products.For example, people need mobility, not cars, or communication, not telephones (Ghafoor et al., 2023;Li et al., 2020;Mont, 2002).Renovation studies like Guerra-Santin et al. (2017) identified the need for a business model to implement and upscale sustainable housing solutions.Balta-Ozkan et al., 2013 andHa et al. (2006) expounded service domains for smart home technologies to meet needs for safety and security, health and welfare, energy saving, and communication within the home.Beyond market factors, Adeyeye (2023) presented an overview of the current regulations and standards for sustainable building performance monitoring and evaluation.Here the logic was reinforced that housing policy could better integrate carrot, stick and tambourine (Hu et al., 2020) services to upscale smart home performance monitoring in the UK, as part of the transition toward net zero homes.
This study aims to (a) investigate how householder perceptions of housing quality, and the drivers and perceived value of home improvements influence their propensity to adopt smart technologies for housing performance monitoring.This was underpinned by models such as the enhanced multi-level research framework for Technology Acceptance and Use (Tamilmani et al., 2021).(b) bridge the theory-to-practice gap by defining the product-as-service approaches for the upscaled implementation of smart home performance monitoring as prioritized by the householders.Studies of this nature are necessary to address the following research gaps (Hargreaves et al., 2013;Li et al., 2018;Rogers et al., 2014;Seminara et al., 2022): (1) The need to establish the extent to which householders may accept or reject smart technology for performance monitoring after they are made knowledgeable of the innovation.
(2) Address the lack of understanding or empirical evidence about how feedback from smart technologies for performance monitoring can be used by householders for their own benefit.This would help to subsequently determine whether performance monitoring will encourage or motivate householders to reduce their levels of consumption and (3) post-occupancy evaluation (POE) remains disparate, and there is a need to join-up the learning and making it more accessible to professionals, policy makers and householders, for better decision making.Thus, in this paper, the types, opportunities, and householder perceptions and affordances for the widescale implementation of smart home performance monitoring are explored.

Housing and policy challenges
The following pragmatic and policy challenges underpin the need for an integrated smart technology / post-occupancy performance monitoring approach as part of aspirations for social, policy, and market transformation in the housing sector (Barnicoat & Danson, 2015;Boardman et al., 2005;Hardill & Olphert, 2012;Ji & Chan, 2020 etc.).The UK's housing policy deployed through Acts, regulations, guidance documents and supporting methodology e.g.SAP, mostly focus on improving new housing quality.Housing turnover is however slow, so net zero ambitions need to better address the legacy housing stock against the backdrop of energy poverty, and other housing inequalities.Over the past decades, there have been many government home improvement initiatives including the recent initiative to subsidize energy bills.Whilst they have achieved partial success, they have not fully acknowledged the behavioural, economic, and technical elements that are purportedly needed to increase the effectiveness of any policy scheme (Elsharkawy & Rutherford, 2018).The short-term efficacy has had minimal impact on long-term housing quality, energy efficiency or the underlying vulnerabilities to external global pricing and environmental trends (Boardman et al., 2005, p. 99).Further, the household demographic, composition, habits, occupancy profiles and patterns all contribute to the current performance gap in homes (Ozarisoy & Altan, 2022).Therefore, policy making is made more complicated by the uncertainties in building performance caused by occupants' behaviour and practices which also affects design decisions for new builds and net zero renovations (Guerra-Santin et al., 2017).The integration of smart technology and post-occupancy performance monitoring would enable these challenges to be addressed.In theory, historic and current occupancy and performance data would be available and accessible to help anticipate future social, economic, and environmental housing resilience scenarios.In practise, decision makers, professionals, and households can be enabled to co-create solutions to address social, economic, and environmental pressures.

Building performance monitoring and evaluation
Post-occupancy evaluation (POE) is the over-arching framework within which Building Performance Monitoring (BPM) and Building performance Evaluation (BPE) are defined.The BS 40101:2022 defines BPM as the gathering of quantitative and qualitative data that characterizes the performance of a building (or separate premises within a building) and the interpretation of these data to draw conclusions regarding specific performance attributes and the overall performance of the building (BSI, 2022).The data derived from BPM underpins BPE i.e. to understand the material performance of buildings according to the type of occupants, as well as how and where resources are used.BPE's value is in ensuring that design methodologies and performance benchmarks are optimized to improve real-world performance.The POE agenda is well-established e.g. in the Royal Institute of British Architects (RIBA) plan of work, but questions about the tangible value to housebuilders, architects, the construction industry, residents, and wider society, about its practical use and efficiency remains (Maslova & Burgess, 2023).BPM remains 'a scanty endeavour of research-oriented academics, rather than being an embedded practice in the building procurement process in the UK' (Durosaiye et al., 2019, p. 347).Housing performance monitoring also requires sector-wide delivery, nudged through policy and regulatory frameworks, as well as an active, willing, proactive, and participating citizenship which are empowered, to engage, trust and adopt in such actions and services (Fakhoury & Aubert, 2015).However, valid questions still need to be addressed such as: who pays and benefits from BPM? Who is responsible if BPM identifies failures and shortcomings?What strategies can be used to achieve improvements if targets are not reached due to occupants' behaviour?(Seminara et al., 2022).

Smart home technologies for performance monitoringopportunities and risks
The number of Internet of Things (IoT) devices worldwide is forecast to almost triple from 9.7 billion in 2020 to more than 29 billion IoT devices by 2030 (Statista, 2022).Global Machine-to-machine (M2M) connections used for automated data transmission and measurement is estimated to grow 2.4-fold, from 6.1 billion in 2018-14.7 billion by 2023; 1.8 M2M connections for each member of the global population (Cisco, 2020).The transitional and transformational approach towards performance monitoring using Big Data, Internet of Things, e-governance, and e-learning is now prevalent in many sectors where the focus on user-centred design is increasingly.Businesses recognize the importance of prioritizing the participative role of the user in 'an iterative design process, as well as the identification of userspecific factors to guide and assess the design' (Eggen et al., 2017, p. 2 ).The housebuilding industry in the UK is also undergoing similar digital and technological transformation (Burgess et al., 2020).Effective user experience research has become an integral part of the housing production cycle and it allows for the testing of design prototypes and for tailoring to ensure the end product is designed to meet user needs (Gothelf & Seiden, 2017).However, this product-oriented focus means that timely user data feedback loops receive less priority.The continuous improvement and transformation of the design and production process thus continues to modernize without improving the quality and performance together with evolving user needs (Maslova & Burgess, 2023).
The 'traditional home' has appliances that are operated locally and manually, usually by flipping a switch or pushing a button.These devices have limited controls and managing energy use can be difficult (Balta-Ozkan et al., 2013).A smart home is a residence equipped with a communications network, linking sensors, domestic appliances, and devices, that can be remotely monitored, accessed, or controlled and which provide services that respond to the needs of its inhabitants (Balta-Ozkan et al., 2014).A typical smart home network comprises a network of smart metres, sensors, home energy management system (HEMS), user interface and communication platforms, and smart appliances (Zhou et al., 2016).The network, through which each of the technological components and information is connected and coordinated, is what distinguishes the smart home from a high tech-equipped residence (Balta-Ozkan et al., 2013).Smart homes data systems are enabled to collect, store, interpret and disseminate two main types of data which according to the BS40101:2022, also underpins the performance monitoring methodology: 1. Relatively static data include (Zhang et al., 2020): a. Building basic information.The information describing the building nature, including an identifier, type, usage, completion time, and so on.This type of data generally requires manual input; however, some existing building monitoring system may provide some related information.b.Building geometry and topology.The information describing the geometry and topology property of the spatial distribution of a building, including area, adjacency, and hierarchy of spaces.This type of data can be obtained from a BIM model.c.Sensor network.The information describing sensor networks in a building to collect energy consumption and environment property, including type, position, accuracy, and collection frequency of each sensor.This type of data can be obtained from a sensor platform; however, the correspondence between sensors and spaces may require some manual work.2. Stream data include the following (Zhang et al., 2020): a. Consumption data.The information describing electricity, water, and gas consumption in different spaces and of different sources.This type of data is automatically and periodically collected by the sensor network and is often stored in a monitoring platform.b.Environment information.The information describing the environment status, including temperature, humidity, CO2 concentration, and so on.These data can be collected by a sensor network.Space occupancy data is often needed to translate this data.
Smart home technologies are increasingly prevalent.In 2020, a compound annual growth rate of 29.5% of the global smart home market was forecasted, with some estimating that a typical family home will contain more than 500 smart devices (Shuhaiber & Mashal, 2019).Aside from personal computers, homes now include technologies that people use for work, entertain as well as the wired and wireless devices that connect to the internet to share information with other people and companies (Brown, 2008).Various domestic systems and appliances are also increasingly connected and converged to handle domestic chores with greater efficiency, and to enable the protection of both property and human lives from danger, including burglary, fire, and flood (Kim & Shcherbakova, 2011).This includes the remote electronic control and management of smart appliances (heaters, air conditioners, washing machines, camera integrated doorbells etc) and realtime access to energy and water usage data.Smart technologies are thus being used by householders or managers to meet pragmatic, hedonistic, regulatory, and altruistic goals.
The user benefits of smart home performance monitoring include energy conservation, utility cost savings, improving quality of life, comfort, convenience, environment friendly, flexibility.The barriers or risks to implementation include lifestyle fit, interoperability, reliability, loss of control, technical complexity, privacy and security risks, administrative difficulties, trust in service provider, utility company or government, acquisition cost, maintenance cost, and knowledge gap (Ji & Chan, 2020).In addition to improving housing quality and performance, smart technologies can be effective for delivering services such as assisted living, energy efficiency, home security or entertainment, as well as facilitating two-way communication between the grid, homes, and any on-site micro-generation (e.g.rooftop solar photovoltaics) (Balta-Ozkan et al., 2013;Balta-Ozkan et al., 2014).Thus, information from smart homes can be used to inform sustainability policies such as public health and wellbeing, energy and water efficiency, security, improved infrastructure, resilience and so on (Ji & Chan, 2020;Wilson et al., 2017).
User benefits are maximized when coupled with products and services (time-of-use tariffs, remote monitoring, efficiency, etc.) (Balta-Ozkan et al., 2013).However, the extensive, and potentially intrusive data collected stored and shared in smart home systems have raised issues of data privacy, and risks to physical security.Despite the benefits, these concerns cannot be ignored as sensors may be used or misused to detect the location of people and objects, or collect data about states (e.g.temperature, energy usage, open windows).If unsecure, the connectivity of domestic devices, appliances and features including anything from washing machines or lighting to a user interface (Balta-Ozkan et al., 2013;Balta-Ozkan et al., 2014;Dong et al., 2018) may provide unwitting remote access to external agents.For this and other reasons, the deployment of smart home technology remains disparate and piecemeal, with little integration of functionality of the hardware and software components.Smart home data are rarely accessible in a singular, wholesome, and usable manner.Data from metres are typically held by utility companies, others held by various providers for security, entertainment, etc.The consequence is that there remains little robust, reliable, and longitudinal evidence to underpin design and performance specifications for buildings.The new Part R of the Building regulations (Building Regulations, 2010b) represents and emerging opportunity which is yet to be fully explored.
Thus, coupling products with services represent opportunities and risks, which can be addressed through local or connected data control.According to Shuhaiber and Mashal (2019), local systems are where different home devices and appliances are connected to and controlled by an internal home gateway through which the occupants can interact with their smart home appliances.Service providers may deliver this service, but the occupants have some control over which data is shared through mobile devices or online platforms.On the other hand, a connected network is where data is automatically collected by external providers from home appliances, sensors, and metres.The data is analysed remotely and communicated to occupants through interactive devices or platforms.Control in both scenarios is achieved by householders' definition and control of what is shared, when, to whom, and how through manual, semi-or fully automated processes.
Householder affordances to adopt smart home performance monitoring must be understood within the context of these opportunities and risks.These underpinned the research design and findings presented in the next sections.

Methods
This study utilizes the hermeneutic research approach to illuminate Gadamerian 'pre-understandings' i.e. exploring perceptions and details within lived experience, and its effects on individuals and society that may be taken for granted (Laverty, 2003;Gadamer, 2010; Figure 1).The methodology is underpinned by a conceptual framework derived from established technology adoption models; from the Technology Acceptance Model -TAM (Davis, 1989) to the extended United Theory of Acceptance and Use of Technology (UTAUT2) (Tamilmani et al., 2021).The methods comprised literature review, survey, interviews, and focus groups until data saturation and validation of the research outcomes.This paper focuses on findings from the survey.Details of the underpinning conceptual framework, and findings from follow-up interviews and focus groups are presented in companion papers.
Technology acceptance factors are well established in literature and include perceived usefulness, perceived ease of use, attitude, perceived cost, perceived risk to privacy, factors involved with social norm and personal norm, like trust in utility company or government, environmental awareness, social contribution including the willing to undertake improvements, innovativeness; as well as facilitating conditional factors, like knowledge, experience, financial capability (Ji & Chan, 2020;Kim & Shcherbakova, 2011).The survey instrument was underpinned by findings from past studies (e.g.Ahn et al., 2016;Balta-Ozkan et al., 2014;Chadwick et al., 2022;Guagnano & Santini, 2020;Ji & Chan, 2020;Li, 2021;Shuhaiber & Mashal, 2019 etc.) and featured the following categories of questions: 1. Housing and household profile 2. Household values and perceptions towards sustainability and technology 3. Factors related to understanding housing quality and performance 4. Factors related to the adoption and usage of smart technologies to support smart home performance monitoring Themes covered perceptions and behaviours, motivations, intent, and factors that inform adoption, and contextual and environmental factors e.g.how technical competencies and trust inform adoption.Where relevant, open text fields were provided to collate additional opinions and viewpoints.The questions used the nominal e.g.categorical Yes/ No, ranked (1-10) or ordinal e.g.Likert scale units of measurement.Statistical methods such as frequency, descriptive, reliability, significance and factor analysis were applied as described in the Findings section.The survey instrument received ethical approval from the relevant University's ethics committee.
Using a social research company, a diverse range of householders from all UK regions were invited to participate in the study.The target was approximately 1000 UK representative sample -66.9 million population +/-3.1% margin of error at 95% confidence interval.The survey was deployed online.The survey started with an introductory video of the key concepts of the study.Data was collected during December 2022 -March 2023 including the telephone surveys to validate the data profile and until a nationally representative quota was achieved.The resulting data was checked, and 972 valid responses were confirmed.The demographic profile of the participants shown in Table 1 provided the starting context for interpreting the data.Analyses were undertaken in SPSS v28.

Household and housing profile
Of the 972 valid responses, 84.1% live as a single family, 9.8% live in a shared house including house of multiple occupancy (HMO), 1.6% live in a household with lodgers, 1.1% in a bedsit or flatlet, 0.5% in purpose-built shared amenities e.g.sheltered housing, 0.4% in temporary housing e.g.hostel, hotel, B&B, and 2.5% in other types of occupancy.Similarly, 22.8% live alone whilst the remainder live with others: 5.1% live with friends or colleagues, 30.1% with partner or spouse only, no children, 4.8% are single parents, 23.6% with partner, spouse and child, children, 3.8% with partner, spouse and child, children, and relatives, 8.8% with extended family (parents, siblings, relatives) and 0.8% in other types of living situation.
The housing profile per regional distribution is shown in Figure 2. 49.4% stated that they live in a town or suburban area, 21.9% in a village, farm or hamlet, and 26.9% live in what they consider a city.Majority live in a house: 19.1% in detached, 28.9% in semidetached and 20.7% in terraced houses.20% live in an apartment, 9.6% in bungalows, and 1.7% in other types of housing.32.4% stated that they own their homes outright, 30.5% own their homes with a mortgage, 1.7% share ownership, 20.1% are renting in the private sector, 14.2% are social tenants, and 1.1% have other types of tenure.
In terms of building services, majority rely on mains electricity and gas supply for their homes.Less than 10 respondents use oil to heat their homes.Only a minority have any form of renewables or water reuse technologies (Figure 3).214 respondents stated that they have water butts for outdoor water use, predominantly in towns and villages compared to the cities.Further, most householders still rely on centralized systems for fuel and power and the use of decentralized renewables, irrespective of the tenure, remain low.For example, only 6.4% said that they have solar PV, the most common form of renewable technologyof which majority were owner occupiers of a type.4.0% have air or ground source heat pumps, here, majority were shared ownership occupiers.

Perceptions of housing quality and home improvements
Housing quality was assessed using 19 questions (Table 2) on a four-point Likert scale: 1 = very poor, 2 = poor, 3 = average good, 4 = very good.The indicators with mean values of 2.8-4.0,implies that the relative importance have been recognized by the respondents.All standard deviation of each category were < 1.0 implying that the discreteness of data is quite close for all categories.Cronbach's α was used to assess the reliability coefficient for the housing quality variables (questions).Cronbach's α value is between 0 and 1, and the closer to 1, the higher reliability in internal consistency (Yu et al., 2019).The Cronbach's α = 0.907, indicates a high level of internal consistency for this scale and this sample.Only the removal of factor 19proximity to local serviceswould result in a marginally lower Cronbach's α.Therefore, all KPIs were retained for the subsequent analysis.Further, with house type as the grouping variable, the Kruskal-Wallis test provided very strong evidence of a difference (p < 0.05) for all housing quality variables except marginally for water efficiency (H2 = 13.9, p = 0.054), and rejected for proximity to local services (H2 = 5.33, p = 0.62).With these preliminary tests confirmed, factor analysis using Principal Axis Factoring (PAF) was undertaken with the 19 questions.Factor Analysis reduces a large set of variables (factors) into a smaller set i.e. a matrix of correlations.This makes it easier to understand the relationship between items in a scale and the underlying factors that the items may have in common.If an item is not related to other items or additional construct need to be explored, then the item communality will be less than 0.40.The higher the communality value, the more the extracted factors explain the variance of the item (Taherdoost et al., 2022;Tavakol & Wetzel, 2020).PAF is an exploratory factor analyses (EFA) approach used to understand the underlying latent constructs that influence how the participants responded to the questions.Whereas Principal component analysis (PCA) is simply used to mathematically shortlist fewer variables (e.g.Loewen & Gonulal, 2015).Factor analysis is best done with a large sample size.Comrey and Lee (1992) considers a sample of 500 very good, and over 1000 excellent, so the threshold was met.Both the Kaiser-Meyer-Olkin Measure of Sampling Adequacy and the Bartlett's Test of Sphericity were passed as the minimum standard on which factor analysis should be conducted.
The housing quality questions examined three factors: quality of the build, efficiency and comfort, space adequacy.Table 2 also summarizes the factor analysis results for housing quality.The first factor, which seems to strongly index quality of the build (fabric) and comfort, had strong loadings on the first eight items.Energy and water efficiency, indoor temperature, humidity, and the ability to control comfort had particularly strong indices.The second factor, which seemed to index site and context factors, had high loadings on the next seven items: noise level, privacy, safety and security, neighbourhood and community, air quality, green space, and natural/ daylight.At loadings < 0.4, quality of the build (fabric) and comfort control had high loadings.Perceptions of quality and comfort is the first factor, with cross-loadings of .47 and .48 on the communalities.It was the reverse for sense of safety and security in the second factor with communality of 0.41 with the first factor.The third factor, which seemed to index design factors, loaded highly on both items, with the type of space having strong communalities with the first and second factors; housing quality and comfort as well as site and context.No loading > 0.4 was indicated for affordability: they returned 0.33 and 0.35 for factors one and two respectively.This was perhaps due to the clarity of the question.Nevertheless, cost and affordability are explored in other questions as discussed in subsequent sections.
Categorical questions with Yes/No responses were used to investigate views about undertaking home improvements.Majority of the respondents (74.5%) would rather make home improvements than relocate to a better one; 87.2% would like to know if their home is performing efficiently; 84% are willing to make home improvements to achieve better efficiency and comfort.Further reflecting previously identified design factors, 66.4% said that they had enough space in and around the house to make improvements to their home.A marginal majority, 51.5% stated that they have the right information and know what to do to make home improvements whilst 49.2% said that it is the responsibility of others e.g.landlords or local authority to improve the quality and efficiency of their homes.Only 39.1% said that they could afford to do home improvements, and only 19.3% said that the government (national and local) are doing enough to improve the quality of homes in the UK.The Cochran's Q Test was appropriate (large, related sample size, dichotomous items; simple random sampling) for examining the null hypothesis.The result determined that there was a statistically significant difference in the proportion of the respondents' views, (χ2(4) = 989.58,p = <.001)' but this is not the complete picture as this test often assumes that the outcome is the same across each category, and that variations are caused by chance.Thus, the pairwise comparison (Table 3) was applied to check correlations such as the correlation between government action and householder drivers to achieve better quality/ higher efficiency home.The findings confirm that housing initiatives by the government drive householder efforts to achieve better housing quality including home improvements.It also shows that agency is not statistically significant in determining responsibilities for, and efforts towards obtaining information and knowledge about home improvement.Further, the knowledge about home performance is not the sole driver for home improvements; cost and affordability, and other factors inform decision making.These findings will be investigated further in the next steps of the study.

Adoption of smart home technologies for performance monitoring
This section explored the technical expertise, awareness, and adoption of other technologies within the home which can inform the propensity to adopt new home technologies.Most respondents (74.4%) broadly considered themselves to be technologically aware, 11.5% consider themselves to be beginners, whilst only 13.9% said that they were experts.Only 22.4% said that they would buy new technologies or devices irrespective of its environmental credentials, 40.5% will adopt technologies after it has been tried and tested, 14.7% are generally reluctant to adopt new technologies, and 37.1% said that they find it difficult to judge the environmental credentials of new technologies including products and devices.
In total, 13.2% said that they have a smart home technology (defined in broad terms).57.7% said that they are aware of it but do not own, 23.8% said that they were somewhat aware and only 5.3% said that they are not aware.The age and education distribution against these findings are shown in Figure 4. 66.4% who owned smart systems were in full time employment compared to 10.9% in part-time employment.Majority of those retired -77.2% are somewhat aware and 32.7% are not aware.44.5% work mainly outside the home whilst 28.9% undertook hybrid working.The Kruskal Wallis test showed no significant distinctions between gender, number in household type, house type, or occupancy profiles.The opposite was the case for age, education, work location and annual household income and social living situation e.g. with friends, partners (with or without children), extended family etc.
The types and range of technologies already owned by the respondents are summarized in Figures 5(a &  b).Home technology service domains were found to include applications for safety and security e.g.intruder alarms and fire alarms, convenience and comfort e.g.robotic devices, healthcare, and welfare -e.g.remote GP consultations, GPS geolocating, childcare, panic alarms, energy efficiency e.g.temperature and lighting control, communicatione.g.online meetings, entertainmentonline videogaming, streaming services, commercee.g.online billing, banking and shopping, and multifunctionalapplying to multiple domains (Balta-Ozkan et al., 2013;Ha et al., 20066).
The Cronbach's α = 0.824 indicates a high level of internal consistency for this scale and this sample.The KR-20 Kuder-Richardson coefficient is sometimes recommended if the data are dichotomous.However, Cronbach's coefficient alpha and KR-20 typically yield the same value (Sijtsma, 2009).Further, with age as the grouping variable, the Kruskal-Wallis test provided very strong evidence of a difference (p < 0.05) for all the devices except smart temperature control, sockets, kitchen appliances, water fittings, entertainment, online services e.g.banking and shopping, remote control, health checks and GPS tracking for persons or belongings.This difference was less apparent when compared by gender.Strong evidence of similarity (p > 0.05) was found for all the devices except for smart lighting, sockets, water fittings, social media, and TV remote.In general, most respondents irrespective of age or gender owned devices that operated within the following service domains: multifunctional e.g.smart phones, personal computing with in-home internet services, and commerce.E.g. online banking and shopping, and ebilling services especially for utilities.TV on Demand and streaming services for entertainment were also highly used.Beyond pragmatic services, the latter suggests that hedonistic benefits (e.g. through gamification) could be useful for promoting smart home technology adoption.
Yes/No responses were again used to gauge the extent to which respondents considered smart homes performance monitoring feasible for them.All levels of awareness considered it expensive to implement.A majority of 70.2% stated that the principal factor for adoption is cost.Followed by 27.1% selected invasion of privacy.Only a minority of 16.3%, 15.2% and 15.6% respectively considered the building age, the adaptability and the disruption to home and lifestyle important factors.The Cochran's Q confirmed that there was statistically significant difference in the proportion of the respondents' views, χ2(4) = 989.58,p = <.001(Table 4).These findings suggest that in addition to cost, other factors such as disruption to household and lifestyles would be considered before adopting smart home performance monitoring.
Respondents were then asked 25 questions across a range of technical, personal, financial, social, and environmental categories.Due to the range of questions, participants could answer yes = 1, no = 2 or maybe = 3.The standard deviation for all categories were < 1.0 implying that the discreteness of data is quite close.The Cronbach's α = 0.948 indicates high internal consistency.Only the removal of factor 25 -'I will adopt if promoted by the media'would result in a marginally lower value.Therefore, all the twenty-five KPIs were retained for subsequent analysis.With the awareness of smart home technologies as the grouping variable, the Kruskal-Wallis test provided very strong evidence of a difference (p ≤ 0.05) for all categories.However, this was not the case when grouped against technical competence.In this instance, all the technical factors including ease of installation, reliability, ease of maintenance, and social factors, showed significant difference (p > 0.05) (Table 5).
Following the protocol as previously described, PAF was again used to prioritize trends.The first factor, which seems to strongly index technical and financial implementation, had strong loadings on the first eight items.As previously indicated, and not surprising in the economic context during which the survey was conducted, cost savings was highly prioritized.Other costbearing technical indices such as energy and water savings, reliability and low maintenance also ranked highly.Some usability indices correlated with this first factor groups e.g.non-interference with other essential services in the home i.e. electricity and internet, promotes data privacy and security including not making the house less secure for its occupants.The affordability of the system also had a strong indication, but comparatively less than the previous categories.The second factor appear to index usability factors.Convenience, compatibility, notifications, and feedback all fall within this factor group.However, the links to technical factors to realize these goals are apparent.The need for an integrated, multifunctional solution was also indexed with strong communality of 0.55.These findings also correlate with findings deduced from Table 4.
The third factor, which seemed to index social factors, loaded highly on the media promoting the uptake, smart technologies improving social image among family, friends, and neighbours, and interestingly, some peer influence, suggesting that uptake will improve if others are seen to be adopting the technology as well.The type of internal space has strong communalities with the first and second factors, housing quality and comfort as well as site and context.No loading > 0.4 was indicated for affordability: 0.33 and 0.35 for factors one and two respectively.With the service and implementation domains established, preferences for smart home BPM service providers were explored.The result determined that there was a statistically significant difference in the proportion of the respondents' views, χ2(6) = 541.77,p = .000.The one-way ANOVA test results (Table 6) indicated p > 0.05 in most instances, which statistically confirms no significant difference between the subgroups: Education, Employment, Annual Household Income, and Region/ County, but not entirely for gender, age, setting (city, town, or village), house type or tenure (see: Ji & Chan, 2020).This considered, 56.3%, 35.2% and 26.7% stated a preference for certified tradesmen e.g.electricians and plumbers, engineers, and designers e.g.architects respectively to undertake the installation of smart home BPM.40.8% favoured any government certified providersnot necessarily building professions.30.9% said housebuilders or providers including housing associations and landlords.19.1% opted for housing or facility managers and surveyors.20.5% elected the DIY option by themselves, family, or friends.A small proportion, 3.9%, selected 'Other' and the comments included: 'a combination of the above', 'it depends', 'the homeowners if it is an existing home, and the housebuilder if a new home', 'not sure … it would have to be a trustworthy government the same goes for building experts', 13 people said no-one, either because they don't think it is necessary, or don't want the surveillance.Some said that they are unsure or don't know (Figure 6).
Further comments and recommendations include: 'A combination of house builders / landlords and energy and water utility companies.Both make continuous huge profits; the government need to incentivise them to do better.Therefore the government should also be involved', 'A combination of housebuilders, government, and home ***owners*** (bearing in mind that many people rent)'.'A combination of the above', 'Depends what it is doing and who benefits', 'New builds should have it', 'probably all the above.there are too many different circumstances that people are in'.A few said that they don't know, and about five commented that no-one should pay because they don't think it is necessary.One person went as far as saying that they 'don't want to be controlled'.Lastly, the participants were asked when they thought that smart homes with performance monitoring technologies would be a reality for every household in the UK.Across all respondents, 21.4% and 11.9% said it will take more than 20 years or never respectively, and only 11.2% said within 5 years.Majority, 55.4% said it would take between 6 and 12 years.Figure 7 shows the findings, in total and by Tenure.

Discussion and conclusion
Most UK householders have little interaction with the regulations and processes that inform the design and construction of their homes.However, performance information and feedback loops can support housing quality improvements, enhance householders' housing experiences and proactively engaging them as co-creators of the net zero transition.With smart home technologies, householder roles could be switched from passive consumers to active resource managers (Ji & Chan, 2020).To succeed, it is necessary to understand technical and socio-psychological factors together (Chadwick et al., 2022;Schot et al., 2016;Sovacool & Furszyfer Del Rio, 2020 etc.) especially where technocratic solutions are being proposed.Therefore, this study investigated the householder affordances for using smart home technologies for continuous UK housing quality improvements through widescale performance monitoring.There were two objectives to this paper: (1) to understand household perceptions about housing quality and performance as the basis to investigate smart home technology affordances and perceived benefits; and (2) to define the service domains with which both widescale adoption and smart performance monitoring benefits can be achieved.
Survey data was used to explore the opportunities of this approach to address the gap between design, regulations and actual housing quality and performance.It was found that householders understand the factorial link between the quality of the build, the efficiency of their homes, and the level of comfort they experience in their homes.Housing quality is a combination of quality of the build, efficient systems, and the adoption of decentralised renewable technologies, although the latter remains low.Efficient systems facilitate energy and water efficiency, indoor temperature, humidity, and the ability to control for these factors.It is noteworthy that the respondents prioritized these tangible performance metrics over perceptual factors e.g.noise and privacy, and contextual factors of location, site, and space.In addition, the propensity to make home improvements to an existing home is found to be higher than relocating to a better performing property.Majority indicated that physical factors e.g.age of the building, space are considered but are not the main limitations to adoption of smart technologies.The cost, the lack of know-how information and housing tenure especially for renters were considered more significant constraints.Other adoption factors include tenure, trust in system functionality, effectiveness and trust in the provider, and administrative factors such as who pays, and who assures the functionality, safety, and security of the system.The perception that the government could do more to address these constraints and improve the performative quality of home in the UK also prevailed.
The data showed willingness to adopt smart home technologies for performance monitoring if risk factors are addressed and benefits can be accrued.The main benefits were to achieve better control of their home environment and use resources in a manner appropriate to their needs, whilst keeping utility costs affordable.Respondents agree that performance monitoring using smart home technologies will generate better awareness of resource consumption in homes and enable householders to adjust consumption behaviours and habits for altruistic and or cost benefit.This affirmed findings from previous research which showed that with smart home technologies, householders can enjoy the convenience of integrated home appliances and systems (Kim et al. 2011), and immediately see the evidenced benefit including cost savings from retrofit or home improvement measures (Haas & Biermayr, 2000).However, it needs to be perceived as convenient, cost-effective, and environmentally beneficial (Balta-Ozkan et al., 2013).Cost was a significant, but not the only adoption factor.Respondents stated other factors like usefulness, functionality, as well as potential disruptions to home and lifestyle.This again supports previous findings that care should be taken with price-based interventions.For example, Nicholls et al. (2017) found that only 26% of households provided with cost-free, off-the-shelf smart energy management devices installed and kept on using them.Smart home technologies may reinforce unsustainable energy consumption patterns in the residential sector, may not be easily accessible by vulnerable consumers, and do little to help the 'energy poor' secure adequate and affordable access to energy at home (Herrero et al., 2018).Therefore, implementation strategies should be designed with care and consider the digital/technical competency divide e.g.among age groups, address agency factors e.g. for leaseholders and renters, and address issues of privacy, safety and trust in government and private providers.
Secondly, the study expounded the benefits of a service-based, rather than product-based approach to maximize the opportunities, and address the risks of smart home performance monitoring It built on Boardman et al. (2005) who proposed policy, social and market transformation to address building performance issues and gaps, and Balta-Ozkan et al. (2013), andHa et al. (2006) who proposed a servicebased approach to deliver home technologies that delivers multidimensional benefit to the householder and external stakeholders.It was found that most respondents currently use a wide range of in-home technologies.Majority of these fall in the multifunctional Table 5. Analysis and factor loadings from principal axis factor analysis with varimax rotation for a three-factor solution for implementation domains (N = 972; 95% confidence interval).domains e.g.smart phones and personal computing devices which predominantly encompass information and communication activities as well as being used for commercee.g.online billing, banking.However, there was also a clear tendency towards the convenience and comfort, and entertainment categories.Therefore, the implementation of smart home performance monitoring should incorporate the information (performance) and communication (dialogue) functionalities.The latter is particularly important to provide advice on what to do to improve performance, but also evidence-as-dialogue to improve service provision, housing standards etc. Further, smart home performance monitoring could satisfy hedonic needs through non-intrusive, intuitive, engaging interfaces for the entire household.Therefore, practically, smart home performance monitoring can be proposed to householders as the following services: (a) commerce (exploiting opportunities for efficiencies and cost savings), (b) convenience and control, (c) entertainment (e.g.gamification), and (d) information and communication domains.The last two domains interact through e.g. through online service and social media interactions.
The findings are consolidated in Figure 8 with recommendations for upscaling smart home performance monitoring as follows: 1. Practicality, cost, and benefit: Ji and Chan (2020) defined 12 KPIs for in-home technology performance.However, building performance monitoring should be understood and proposed from two angles: (a) functional/ pragmatic i.e. for the purpose of providing information for the understanding, management, and optimization of building performance, and (b) dialogue based i.e. delivering information, participation, and feedback service for householders.As found in this study, most householders already engage with providers for various information-based services e.g. for utilities, banking, healthcare, social and personal media.Privacy, safety, security risks exist for these services, to a greater or lesser extent as sharing building performance information.Therefore, it should not be presumed that householders would be unwilling to share building performance data.The question that needs to be addressed is 'what is the gain'?How can performance monitoring transcend the pragmatic domain to the service/ dialogue/ benefit domain?In this regard, the benefit of a smart home performance monitoring to the householders becomes paramount.The findings showed that respondents prioritized the capacity of such system to deliver cost savings.This aligns with the need for such system to deliver prompts, and behavioural feedback, and troubleshoot and indicate faults and maintenance issues before they become costly e.g.leaks.Since one purpose of the smart home is to improve quality of life rather than complicate it, there is also a need for a type of smart home system that is outwardly intuitive and easy to use.All systems should also be technically functional and reliable, require little or no maintenance, be interoperable.However, with the right policy, commercial and legal sensibilities, these technical and administrative issues are easily scalable and resolvable by competent third-party smart home providers.2. Policy initiatives for the implementation of smart performance monitoring should target three areas: information, finance, and tenure.These are essential to improve the agency and participation of households in the government net zero homes ambitions.Knowledgeable and empowered householders are more likely to demand and drive market change in the housing sector.National and local authorities could work together, to enhance existing services into coherent, visible hubs for authoritative information for sustainable design, retrofit and adoption of renewables.Like the exemplar homes in the planning portal, interactive exemplars of smart home performance monitoring could be showcased to home buyers, renters, or renovators.The government through these physical or virtual hubs could signpost opportunities and best practise, showcase approved installers, and facilitate incentives and/or rebates to address cost barriers.Lastly, policy initiatives targeted at improving building quality and performance in the rental sector is often overlooked.Addressing this gap would be crucial to the participation of leaseholders and renters in performance monitoring, smart or otherwise.3. Smart homes performance monitoring as a Service: All smart home systems need to be affordable for the household, and effectively integrated in new and existing homes.Cost was consistently raised as a barrier to adoption.Therefore, a service innovation approach would be essential for uptake.
Most new homes come pre-installed with internet, telephone and TV infrastructure as facilitated by the new Part R of the Building Regulations (Building Regulations, 2010b).Therefore, it makes sense to include smart performance monitoring as part of this regulation for new or retrofitted housing, consolidating Part Q (Building Regulations, 2010a) for enhanced security.This could mobilize the emerging market, create jobs and give householders options to engage with a service-based model akin to leasing or purchasing a TV, broadband, or telephone subscription service.This approach will also address the concerns about the technical expertise and know-how to purchase, install and maintain the right systems safely and securely.The user's confidence would improve only if a minimum level of expertise is required to use, manage, and troubleshoot the system.4. Housing sector transformation: The smart home performance monitoring -technology and services should be well integrated into the design, use, and general sense of home.Technology that does not fit in with the surroundings, pre-existing norms or know-how is less likely to appeal to the homeowner and consumer (Stringer et al., 2006) and may contribute to a feeling of being 'out of control' (Balta-Ozkan et al., 2013;Stringer et al., 2006).Smart home technology should therefore 'fit in' physically in terms by being suitably installed and integrated into the structure of the residence and its contents.To some degree, it should also fit in aesthetically in terms of the look, shape, and colour of the various components (Balta-Ozkan et al., 2013).The concept of plug-nplay should also be explored as smart home performance monitoring should evolve to meet the evolving needs, demands and preferences of its occupant users.To avoid becoming redundant, it should be feasible to upgrade hardware and software components in an ever-changing landscape of the technology industry (Balta-Ozkan et al., 2013).
In conclusion, this paper bridges the theoretical and practical implementation gap for the widescale deployment of smart home technologies for performance monitoring.It found that householders are broadly willing to adopt smart home performance monitoring in exchange for a service or benefit.Service innovation models should be used to address cost and technical issues e.g.system complexity, maintenance etc.The service domains with which such a scheme could be established are also defined.There are limitations to these findings, including the extent to which survey instruments are effective for fully understanding perceptual factors.The study reinforced perceptual risks to data privacy and raises issues of consent.Other factors like the impact on tenure e.g.owns versus rent also need to be further investigated.These limitations and risk factors will be explored in the next stages of the study.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Figure 3 .
Figure 3. Building services profile (Alluvial diagram size in descending order).

Figure 4 .
Figure 4. Awareness of smart home technologies indicated by age and education.

Figure 6 .
Figure 6.Responses to who should pay for in-home smart performance monitoring systems.

Figure 7 .
Figure 7. Respondents forecasting the adoption of smart home technologies in UK homes (by tenure).

Figure 8 .
Figure 8. Service domains and implementation strategies for smart home building performance monitoring.

Table 2 .
Analysis and factor loadings from principal axis factor analysis with varimax rotation for a three-factor solution for housing quality questions (N = 972).
*Loadings < .40 are omitted.Kruskal Wallis Test grouping variable: House type.BUILDING RESEARCH & INFORMATION

Table 3 .
Pairwise comparisons for non-statistically different home improvement factors.The government / local authorities are doing enough to improve the quality of homes -I will move rather than improve the quality of my home It is the landlords/ local authority responsibility to improve the quality and efficiency of my home -I have the right information and know what to do Each row tests the null hypothesis that the Sample 1 and Sample 2 distributions are the same.Asymptotic significances (2-sided tests) are displayed.The significance level is .050. a Significance values have been adjusted by the Bonferroni correction for multiple tests.

Table 4 .
Pairwise comparisons for non-statistically different home improvement factors.Each row tests the null hypothesis that the Sample 1 and Sample 2 distributions are the same.Asymptotic significances (2-sided tests) are displayed.The significance level is .050. a Significance values have been adjusted by the Bonferroni correction for multiple tests.

Table 6 .
Difference analysis for service provider between the sub-groups and by demographical variable (n = 972).