Keywords

1 Introduction

Studying the home is a difficult endeavor, let alone kitchen practices, as meaning is created through informal agreements, tacit knowledge and embodied interactions [36]. The majority of people’s actions within the home occur in an unstructured manner and are not just part of routines [11]. Notwithstanding, smart home research either focuses automating routines or improving performance [10].

Other design strategies should be taken into account when designing experiences for the home, and in particular for todays kitchens. Experiences cannot be boiled down to human-system interactions, but stretch from interactions with other people, content, dumb and smart objects to a multitude of situational contexts [51]. Therefore, it is important that technology design is informed by understanding the needs of people, activities they perform in the kitchen, and the situations taking place in household kitchens [38]. Further, design choices should enhance (and not distract) traditional living, and match (instead of conflicting with) the original interactions [8]. Joint HCI and cooking literature typically limit their focus to one type of interaction of the cooking experience, solely focusing on either content interactions, object interactions or social interactions. In an IoT context, however, interactions with technology have become multidimensional and multimodal and woven into everyday interactions [52]. Hence, the authors of this paper strongly support a holistic approach applied in successful user experience design [40], taking into account all interactions that might impact or result from the envisioned human-system interactions. Therefore, the empirical contribution of this study lies in the holistic assessment of the kitchen ecosystem from a user-centered approach. Moreover, this study also provides a practical contribution for designers and engineers by means of design requirements.

2 Related Work

2.1 Experience Design

Experience design is in essence about problem setting and problem solving, which strives to design the right functionality and the right concept [7, 12]. All product components, such as its functionalities, content, interactions, interface and tangible objects should be aligned with the foreseen experience. Therefore, Hassenzahl suggests to first ‘consider experience before products’ [22:76]. Moreover, it is argued that researchers should relate empirical insights in the context-of-use with theories and models [4, 22].

Domestic cooking experiences are complex experiences consisting of a multitude of interactions, such as interactions with food, with kitchen appliances, with recipes, with other household members, with devices and tasks unrelated to cooking, etc. By considering the holistic context-of-use and all these interactions, designers are able to make more informed decisions regarding future kitchen products and services.

In the next section, we will give a short overview of the main components studied in HCI research regarding domestic cooking experiences.

2.2 HCI in the Domestic Cooking Experience

Efficiency Versus Experiential Quality

Since the advent of smart homes, HCI has also moved into the kitchen. Ethnography studies suggest that kitchens are not only goal-oriented places [5], but rather places where common everyday experiences take place [36]. However, the innovations that have entered modern kitchens or that are presented at HCI conferences merely focus on automating habits and assisting household cooks. For example, KitchenSense, translates clusters of routines into personalized information on tasks (e.g., safety alarm on cooking temperature, cooking technique instructions) [26], and the visual prompting system of Ikeda et al. [24] assists elderly in multi-step tasks in the kitchen.

The strong focus on efficiency over experiential quality has led to a wide arsenal of kitchen appliances, such as microwaves, deep fryers and food processors, which resulted in faster, but not necessarily better food [2]. Grimes and Harper [17] also found in their review of food research in HCI that technologies are informed by a corrective approach and are thus designed to fix problems that individuals are thought to have with food. Therefore, Bell, Blythe and Sengers see potential for HCI research to resolve the borders of design spaces by criticizing the assumptions that underlie the design of everyday objects [2]. In their research they defamiliarized the home through the use of ethnography of English and Asian homes, and a cultural history of American kitchen technology and translated it into twelve design recommendations. None of the 12 design statements values efficiency, whereas two statements are related to hedonic experiences (i.e., creating new experiences and supporting serious play), and four shed light on the needs and motivations of the user. The remaining six statements take the broader socio-cultural and situational context into account. Hence, recent efforts in HCI research encourage ‘celebratory technologies’, that complement existing positive interactions or with food instead of focusing on problematic interactions that technology might ‘solve’ [17]. One such technology is the Living Cookbook [47], by means of which people can share live cooking experiences with others or teaching how to cook.

In order to avoid that celebratory technology ideas turns an already positive experience into a negative experience, however, the authors emphasize the importance of understanding the behavior being addressed [17].

Social Interaction in the Domestic Cooking Environment

Kitchens are inhabited by different household members, each with different motivations for being there. One could see the kitchen as the ultimate play zone, whereas others perceive it as a place to rest, to connect with others or to improve their cooking skills. The social context of living together with others impacts cooking experiences and the meaning people attach to it. Multiple studies found that living together with other people structures cooking practices, because sharing homemade meals serves as a symbol of their family life [53, 54]. In contrast, people living alone, such as widows, consider cooking rather as a chore and are less willing to invest time and effort to cooking [44].

In a qualitative study that explored hedonic eating experiences of 16 people, was found that social interactions positively contribute to hedonic eating [33]. In particular the presence of others and the eating activity serve as a venue to other social activities (e.g., talking about food, sharing thoughts, preparing meal together).

A video observation study on the gestural interactions during family cooking acknowledged the importance of social context for gestural interactions [36]. Whether people preferred verbal or gestural interactions, depends on the social characteristics of the situation, such as age, ethnicity, type of relationship, and food related knowledge and skills. However, it is this presence of multiple people packed in one space that inhibits the successful recognition of gestures and different people by gestural sensors [38]. With regard to the cooking stage, it is argued that cooking food is a more individual activity than food preparing, with the latest holding the opportunity to be a ‘powerful bonding mechanism’, as intimate social interactions take place while preparing ingredients [32:282, 35]. Not only the cooking stage, but far more the physical context of kitchens facilitates socializing among significant others [32] and should be considered as an integral part of the cooking experience [21]. The physical activity of preparing food in combination with intimate conversations are even considered ‘kitchen therapy’ [32:282].

Interaction Modalities

Kitchens are multisensory spaces that cater to the experience of processing and combining ingredients during which odors and flavors emerge. However, systems that are present in today’s kitchens do not respond to these type of experiences. Interfaces of ovens, scales, hobs, and even novel innovations that are presented at HCI conferences are primarily graphical user interfaces with touch interaction modalities [13, 20, 27 e.g., 42]. Tangible technologies mostly facilitate pragmatic experiences, and to a lesser extent social and hedonic experiences [30, 36, see: 49]. Based on their ethnographic study on family cooking, Nansen, Davis, Vetere et al. [36] see potential for tangible technologies to support special cooking occasions and for (non-verbal) system-mediated communication. Furthermore, they suggest that especially younger cooks will benefit from grip-sensitive sensors and adults rather from speech input. This was also found in a preliminary evaluation study of an interactive recipe system [35]. Its tangible tags that visualize the available ingredients facilitate social interaction and decision-making.

An observational pilot study towards the development of the MimiCook system, in which recipes are embodied in the kitchen counter, discovered that people misread the amount of ingredients or overlook a step when reading step-by-step instructions of a recipe [42]. This is a common difficulty with graphical user interfaces, and hence also traditional paper recipes, due to a decoupling of the perception and action space [43]. Tangible interfaces that unify input and output can overcome this difficulty [43]. Once the technology was tested, researchers discovered that the display time of instructions is critical, as well as the interpretation of instructions such as “mix well” [42]. Moreover, the study showed that people were confused what measuring tools or utensils to use [42].

The Broader Situational Context of Cooking

Situational context has shown to impact user requirements regarding design and features of kitchen technology [14]. To our knowledge only one study evaluated the potential of interaction modalities of a cooking assistant prototype for different situations, such as cooking new versus known dishes, busy cooking (i.e., while multitasking) and mobile cooking (i.e., cooking outside of anybody’s home) [50]. This study found that how features are perceived, largely depend on context and age. Social context raise as an important not a priori defined context as one to impact preferences of interaction modalities. In situations where other people are present during the cooking process, for example, speech input is considered socially inappropriate [50]. The cooking assistant prototype could also automatically adapt to the context of use, such as postponing audio messages until a conversation break, which was highly appreciated by the participants [50]. However, turning an ubiquitous system into blind and deaf mode was not yet implemented, although much sought after [50].

2.3 Towards Situated Cooking Action Observations

While several studies have been conducted regarding user experience of domestic cooking technologies, these are limited to either the investigation of one specific component of the cooking environment and/or the evaluation of a specific technology. A study mapping the domestic cooking ecosystem can provide an interesting step back in order to define both problems and current positive aspects of the domestic cooking experience. This can provide a meaningful framework for the conceptual design stages of kitchen technology and thus design by understanding the activity [36]. While several studies have conducted research aiming to understand the domestic cooking activity, these were either a) a small discussed step in a technology evaluation oriented paper [42] b) not conducted in a domestic environment [42] and/or c) aimed towards a subcomponent of the whole cooking experience [38]. The present manuscript goes beyond the state of the art by investigating the domestic cooking experience in a holistic manner.

Moreover, this manuscript goes beyond the state of the art through its methodology. Previous research aiming to get insight in the domestic cooking experience, either asked participants to gather visual data of themselves that was later used during in-depth interviews [55], conducted (video-) observation of the cooking activity at home with a researcher present [55] and/or created an artificial domestic cooking experience outside the homes of the participants while being observed by a researcher [42, 50]. This results in representations and versions of events, produced by the researcher and/or the participant [55]. More specifically, these are either reconstructions and/or projections of actions. According to Suchman, however, actual in situ action needs to be distinguished from projected actions (i.e., a priori rational plans) and reconstructions of courses of actions (i.e., post-hoc interpretation of actions) [45]. Actions depend upon the action’s circumstances and comprises of ad hoc responses to contingencies of particular situations and the actions of others [45]. This means that the investigation of domestic-cooking experiences should occur in its authentic environment, not only referring to the physical environment, but also to the presence of other household members during this activity and other relevant events or interactions that influence actions of the household cooks. A study of situated action of the domestic cooking experience is not only important to gain an ecologically more valid view on this experience, but is also important in order to gain insight in which situations influence actions and how these influence action. This is important in HCI design as machines do not have the same resources as humans to interpret situations of actions. Successful interaction between people is however, the result of mutual understanding of situation and through the detection, repair and exploitation of possible differences in understanding [45]. Successful interaction between ‘smart’ technology and humans should thus also thrive for the same, as a machine’s insensitivity to situations have previously resulted in negative user experiences [45, 50].

By using video cameras in the home context, without the presence of a researcher and without actively asking participants to shoot content, we aim towards a more ecologically valid view of the domestic cooking experience, its actions and its situations of actions. The present manuscript discusses preliminary results.

3 Methodology

The aim of this study was to gain a better understanding of cooking experiences in today’s Western kitchens. For this purpose, in-depth interviews have been conducted with 16 people facilitated by retrospective think-aloud protocol (RTA). Prior to the interview, the interviewees participated in in-situ video-observations that have been conducted in their kitchens during two cooking moments. The observations (only video, no audio) took place on a weekday and a weekend day. The observations lasted approximately 1–3 h per cooking moment. The research design has been approved by the faculty’s ethical committee.

3.1 Video Observations

One week before the in-depth interview, two Nest cameras were installed in the kitchen of the participant. One camera (cooking camera) was installed nearby the cooking hob that captures all interactions with the hob, hood, and pots. The other camera (context camera) captures the broader physical context of the kitchen, focusing on the back of the cook and the most frequently used passages (see Fig. 1). During the camera setup, participants were asked when they would be cooking during the next week and weekend. Accordingly, only data is collected during these agreed-upon cooking moments using the calendar functionality of Nest. A Nest camera can be remotely switched on and off, and family members are notified of its power state through a green led on the camera. Before data collection starts, all adult family members are provided with a consent that informs them about the purpose of the study and the data that is going to be collected. This informed consent requests their voluntary participation in the observational study and a follow-up interview with the focal participant afterwards. In case the participating family has one or more children younger than 16 y/o, parents also have to give their consent to videotape the children.

Fig. 1.
figure 1

Video observation set-up

Full size image

The video observations served as probing material of the Retrospective Think-Aloud Protocol. During the qualitative data analyses of the interviews, the videos recordings also served as additional contextual information.

3.2 Retrospective Think-Aloud Protocol

In the retrospective think aloud (RTA) protocol participants verbalized their thoughts while watching the video recordings of their cooking. In contrast to concurrent think-aloud (where participants think out loud while working on a task) this method yields more explanatory insights, as no cognitive effort is imposed on participants while undertaking the cooking task [19]. Playback speed was altered to four times the original speed, resulting in playback durations from 40” to 1h40” for the recordings of the combined cooking moments. Video recordings were paused if needed. Moreover, if particular interactions or contextual characteristics were not coined by the end of the RTA, a semi-structured topic guide was followed covering questions, such as ‘Is there a division of roles between you and your partner in the kitchen?’, ‘What would you like to change in your kitchen?’, ‘What makes you insecure while cooking?’, ‘What is the meaning of your kitchen?’, ‘How would you describe yourself as a cook?

3.3 Data Analysis

The in-depth interviews are transcribed and coded in NVivo 12. Data coding proceeded following the analytical process of Grounded Theory [16], which consists of multiple waves of open, axial and selective coding. Preliminary results in the open coding stage revealed interactions with ingredients, kitchen utensils, kitchen technology, kitchen technology interfaces, other people, non-related kitchen technology and content. Additionally, time-related variables, space-related variables and dish type resulted from open coding. By means of axial coding, these open codes were subdivided into the following categories: social context, information and task context, physical context, temporal context and situational context. In the selective coding phase, the authors decided to select the situational context as core category, as it serves as a lens for people to motivate and perceive people’s cooking experiences. In the results section different situational contexts are delineated and how these impact the perceived interactions with objects, cooking appliances, content and other people in the household, across the different cooking stages.

3.4 Participants

Participants were recruited from a profiling survey that has been conducted as part of the same research project. While a full description of this profiling is beyond the scope of this paper, in brief, it was based on a k-means clustering of survey questions relating cooking motivation and behavior. It resulted in three profiles of increasing levels of motivation to cook and action cooking behavior (i.e.: 1) persons least motivated to cook, 2) those struggling to find a balance between cooking and other pressing activities and, 3) those with a motivated and creative cooking profile). These three cooking profiles were also present among the 16 participating cooks in this study. Ten participants are female, and the average age is 33.69 years (SD = 7.74) with the youngest 27 and the oldest participant 52 y/o. Six participants live together with their partner, another six with their partner and children, and three people live alone. The least motivated cooking profile with the lowest cooking frequency is represented by three people, the family cook that struggles with a healthy work-life balance is represented by six participants and another seven people fit the most motivated and creative cooking profile. The names of participants, family members and identity-sensitive video data are anonymized.

4 Results

4.1 Physical Context

Half of the observed households have a semi-open plan kitchen (i.e., the kitchen is separated from the living room through a kitchen island or low wall). Two open-plan kitchens (i.e., kitchen and living room are perceived as one room) and three closed-plan kitchens (i.e., the kitchen is located in a closed room separated from the living room) set also the scene for the observations. Participants who are delighted about the kitchen’s physicality, most of all they praised the kitchen’s atmosphere, i.e., soundscape and lighting. Whereas some prefer the radio to be on, the other prefers its silence. Also, its lighting seems to be crucial for the positive evaluation of a kitchen. Further, cooks positively perceived their kitchen’s ergonomics, openness, and size. In their kitchens everything was within easy reach and open-plan kitchens served as a venue to more social contact. With regard to kitchen size, they are satisfied about the large surface or in case of a small kitchen, it forces them to be creative with the available resources. A man who rents a small house with a closed-plan kitchen describes this tradeoff between lack of space and creativity:

Look, right now, over there is not enough space. That kettle was on the floor. I will put it back on the hob. It is really a matter of a lack of space, but I like to cook in a small kitchen. It challenges me to be creative though it is frustrating. It structures my cooking and forces me to create less chaos.

(P12, man, 31 y/o)

More often, on the other hand, household cooks are bothered with the physical characteristics of their kitchen, with atmosphere (i.e., gloomy lighting, environmental noise, and size (i.e., lack of space) as the most disturbing elements. A kitchen that always turns out to be messy, that forces the cook to do the dishes in between cooking, or that offers the cook limited space to prepare ingredients are often-mentioned small kitchen annoyances.

Objects

Kitchen utensils, pots, pans and ingredients are the main objects where household cooks interact with in the domestic cooking environment related to cooking. Smartphones are a non-cooking related object that is present among almost all participants and often used during ‘dead’ moments. Pots that need to be stirred and require attention, such as saucepans, are put on the front of the hob. High pots with contents requiring less attention are put in the back.

When cooking alone, cooks get a positive experience from efficiency-oriented practices, such as optimizing utensil use in order to limit dishes. Related to this, negative experiences are created when size of pots and pans needed for preparing a dish are underestimated, and one has to change pans resulting in more dishes.

Kitchen Appliances

Main kitchen technology that participants interact with are the oven, the hood, the hob and the fridge. The hood is turned on if the stove is turned on or when the pots are getting warm, because of its integrated lights. The hood is not always turned on and two main reasons can be distinguished for this: either it is unintentional (i.e., forgotten) or intentional, as the hood makes too much noise. When turned on, either the middle or highest hood level is used: The lowest hood level is almost never used, as ‘it is so quit, it can’t possibly extract fumes’ Only a minority leaves the cooking hood on after cooking, which is related to the quality of the hood (i.e., too noisy).

The hob is –besides actual cooking- also used as a source of warmth, keeping food warm while eating. The residual heat is also often used for cleaning burnt remains. Mixing soup is also an activity that occurs on the hob.

Remarkably, cooks who were used to physical button as interface with the kitchen technology prefer this over their current touch interfaces. They are perceived as more easy, fast and reliable. Also, implicit interactions such as in the induction system, are sometimes considered frustrating. For instance, when the stove is turned off as a result of over boiling water.

One participant stated that she loved to cook with the oven, because ‘you don’t have to keep an eye on the oven all the time and you don’t have to stay in the kitchen’.

4.2 Social Context

A difference between active and passive household members can be made. More specifically, active household members refer to household members involved in the cooking process. Passive household members refer to members present in the physical context around the cooking experience, not involved in the cooking process, influencing the domestic cooking experience.

Only three participants prepared a meal alone during both cooking moments and only two were supported by their partner during both sessions. However, cooking alone does not mean that no one was nearby; they were accompanied by their partner, children, both, cohousing friends or a household help. Hence, it is noteworthy that people who cook alone often talk with others who are present in the kitchen or in the semi-attached living room. The observations show that couples tend to cook together more often (all but one participant cooked at least once together) and families with children tend to divide chores between both parents with one being occupied with dinner preparation and the other mostly occupied with the children. Only three couples cooked together or with their children, the remaining eight family cooks prepared a meal alone during one or both cooking moments.

All but one participant mentioned the social context as determining for their cooking experience in either a good or bad way. Especially in terms of arousal, such as talking, intimate affection, and the pleasant feeling of having someone around. One participant also mentioned the increase in efficiency as and advantage of cooking together.

It’s not like we need each other to cook. We don’t wait till the other is home to start preparing dinner. It is more fun, I mean, being together in the kitchen is the first thing that happens if we turn back from work. That is our main reason, to have a little chat. (P07, man, 27 y/o)

Contrary, the social presence of others or even cooking together leads to a plethora of frustrations and conflicting processes among 8 participants. Especially the presence of children, which negatively impacts the feeling of control, leads to unforeseen conditions, and makes it difficult to divide attention between cooking and children. When cooking together with others, people mentioned conflicting priorities (slow vs fast, precise vs loose, …) or processes (more dishes, getting in each other’s way, no multitasking, already in use utensils or cooking zones …). Moreover, cooking together appears more chaotic and inevitably it results in waiting for one another. One woman (52 y/o) also mentioned that cooking with other people than her partner makes her insecure about her cooking skills. Another woman with two children of 1 and 3 years old explains why she strives to do the dishes together with her partner in the weekends:

Mostly my daughter grabs on to my leg and then I can’t do much anymore. Now (referring to the video observation) both kids are in the storage room opening and closing the door over and over again, but I give them a hand slap and tell them to stop it.

(P05, woman, 33 y/o)

Another woman describes why she prefers to cook alone instead of together with her partner:

I have the feeling that when I cook alone on weekdays, it occurs more efficiently if no one disturbs me: I immediately clean everything. But, if we cook together, the dishes start to pile up and everything gets everywhere. During those moments I think ‘that doesn’t belong over there’. Maybe I think I am more efficient if I do it all by myself?

(P02, woman, 27 y/o)

4.3 Information and Task Context

Seven participants cooked one or both cooking moments with a recipe. While not all cooks cook with a recipe, those who do, follow it closely. Insecurities raise as a result of non-detailed description of quantities, and not knowing how something should look like and when something is done.

In this study, cooks explained that they prepared new recipes more often for guests than for their family. When cooking for guests, they prepare more elaborate meals, which provides more pressure, but is also worth it according to three people. Only one observed cooking moment was for guests (single woman, cooking for a friend) and all other participants cooked both observed moments for their family.or in case of singles for. Remarkably, participants evaluated the preparation of a family meal based on other parameters than a meal for guests. When describing the cooking process for guests it is both evaluated in terms of its result and the overall experience, whereas family meal cooking is described in terms of its result, i.e., goal-oriented cooking. A single woman who cooked dinner for a friend explains why her attitude during the think-aloud was rather negative:

It seems like I don’t like cooking, but it’s just because I’m under time pressure. You are cooking for someone else. So, it has to be good, because you want her to like it. The day before it didn’t really matter, because I had to eat it myself.

(P01, woman, 27 y/o)

Almost everyone (n = 15, f = 22) prepared a routine dish at least once, whereas eight cooks prepared a new dish with a recipe (f = 7, n = 6) or improvised a meal from scratch (n = 2, without recipe support). People oftentimes explained their choice for familiar dishes in terms of the needs and restrictions of family members, or depending on the week of the day (weekdays vs weekend). These household cooks prepare a dish that everyone likes, which are mostly simple and fast dishes. Furthermore, the narrow preferences of children and food intolerance negatively impact meal variation. The most prepared meal during the observations is a typical Flemish meal, consisting of sausages, apple compote and mashed potatoes (WAP).

We prepared WAP, but we used frozen mashed potatoes, since there were no potatoes left. On Sunday evenings it is mostly a quite simple meal, which is not an exception over here.

(P07, man, 27 y/o)

4.4 Temporal Context

All participants cooked once on a weekday and once during the weekend as prescribed in the research design. The majority prepared a familiar dish on weekdays (n = 13), whereas a more equitable distribution of dish type can be noticed during the weekend (familiar dishes, n = 9 - new dishes, n = 6). This is indicative for the close relationship between temporal and task context. If participants considered the temporal context of their cooking experience, it is mostly described in terms of time pressure and routines. Time pressure results especially from the cooking event as part of a chain of events, such as after grocery shopping and before sports activities, but can also result from the cooking experience itself. One woman who works fulltime and has two children describes her personal experience with cooking on weekdays.

P03: There is a big difference between cooking on weekdays and in the weekend. I don’t feel at ease when I don’t have enough time to cook. It’s like going for a run to relax, but you have to be back within half an hour!

Interviewer: Why is it similar?

P03: Then I don’t cook, or I opt for something very easy

Interviewer: When do you not feel at ease to cook?

P03: If I arrive late home or if we have to go somewhere. (woman, 32 y/o)

Another single participant who cooks for a friend feels agitated while cooking, due to unforeseen circumstances. She has to switch pots and the cooking process proceeds faster than expected.

I used a small pan to fry the onion, but then I realized I need a bigger one to add the other ingredients. I switched pans, while I thought ‘now I have more dishes’. Now I add the olive oil, which warmed up quickly and I get even more agitated.

(woman, 27 y/o)

5 Conclusion and Discussion

Based on the preliminary analysis of the in-depth interviews, we can deduct the following situations:

  • cooking for a guest (external household) or cooking for family/cooking or for yourself (internal household)

  • cooking routine dishes or cooking new dishes

  • cooking on weekdays or cooking on weekend days

  • cooking alone or cooking together

These situations impact requirements regarding development and design of kitchen technology and applications. Situational context impacts temporal context, information and task context and how the physical context is perceived. Temporal context, information and task context and physical context in turn impact the interaction with objects, content, kitchen technology and other household members present in the direct domestic cooking environment. When aiming to develop IoT technology for the kitchen, a system should be able to detect these situations and adjust its behavior depending on these situations.

When cooking during weekdays, for instance, temporal context takes the upper hand: people are often under time pressure because of other activities before or after the cooking activity. This results in people resorting to routine dishes and thus cooking without a recipe (information and task context). During weekdays, children are present during these cooking moments, possibly distracting the cook (social context). Main frustrations are thus related to not being able to finish the meal in time or that separate subcomponents of the dish are not finished in time and being distracted while cooking. A smart kitchen system that can support households in reducing time pressure during weekdays, both during preparation including grocery shopping and cooking phases would be of added value. Even more if it allows the cook to being distracted, without having to sacrifice time or dish quality. This system should, however, allow for input of the cook -and not rely on recipes of the system- as during these time pressured moments one falls back on own recipes.

When cooking during weekend days there is less time pressure and more opportunity for hedonic cooking. More specifically, trying out new dishes and learning new cooking techniques (task context). A smart kitchen system should thus also be able to inspire cooks before (e.g. system-suggested recipes that require more time/skill so the cook can learn new dishes) and during the cooking process (e.g. trying out a new slicing technique when cooking in the weekend).

Cooking alone is often considered more efficient than cooking together. This is mainly caused by differences in speed of separate cooking processes, different ways of following a recipe (e.g. one wants to follow the recipe rigorously while the other likes to follow it more freely, changing several components/information and task context) or lack of space (physical context), all resulting in frustration. However, people like to cook together as it is a ‘social moment’, a moment to talk about how the day went, to show some affection, etc. Here, we thus see a clash between temporal context elements (efficiency) and social context elements (social interaction with family members). Cooking together for most of the time, means that different participants of the household participate in different steps of the cooking process, such as preparing (cutting vegetables) and actual cooking. In these interactions, it is also noticeable that there is always one ‘chef’ in this cooking process and one ‘sous-chef’. A smart kitchen technology that supports this cooking together process in terms of several stages of the cooking process and several roles, might reduce frustrations during the ‘together cooking’ situation and make it as efficient as cooking alone. It might also offer the opportunity for the less experiences cooks in the household to learn from the household ‘chefs’. When cooking alone and for oneself, efficiency regarding time (temporal context) and object-use (i.e., reducing dishes) are more important.

When cooking a dish for the first time, the information and information context and temporal context variables are most important. More uncertainties exist and more interaction with content will be required (information context) A technology focusing on not only user-content interaction (user following recipe) but also for system-user interaction, in the form of feedback content to address these uncertainties might be an interesting opportunity. Cooking a dish for the first time also influences temporal context, as one typically chooses for a moment when more time is available to experiment. If there is a clash with temporal context, a negative experience will be generated. When cooking routine dishes, there are little frictions in the cooking experience. However, the choice of routine dishes is often out of comfort considerations. More specifically, a sense of control regarding time (i.e., one knows approximately how long it takes to make the dish and outcome and outcome (i.e., one knows family members will like it). Kitchen technology enhancing the experience, providing small alterations or suggestions for similar dishes regarding time of preparation and taste might be an interesting opportunity.

While this manuscript only discusses preliminary results of an in situ video-observational study combined with in-depth interview research using retrospective think-aloud, we can state that situational context is an important starting point when developing new technologies for the domestic cooking environment. It will influence temporal context, social context, physical context and information and task context. These will in turn influence requirements regarding kitchen technologies and applications. However, other objects present in the direct context-of-use, other users (passive or active) and other content, will also influence the user experience of cook with a kitchen technology. This study has contributed to the state-of-the-art, as – to our knowledge- no studies have systematically analyzed domestic cooking experiences using a situated action approach. One study had previously defined relevant situations to take into account when evaluating technology features [50]. This was, however, not based on previous analysis of possible situational contexts. In this study, the contexts were created and impact of the context on feature preferences were examined. In line with their assumptions, we have defined the ‘cooking a routine dish versus cooking a new dish’ and the ‘cooking together versus cooking alone’ situation.

6 Further Research and Limitations

These results need to be further validated, as for instance, in only one occasion did the ‘cooking for an external household guest’ situation occur. Moreover, relationships between the different broader (non-tangible) contexts and elements within the direct context-of-use (i.e., objects, kitchen appliances, content and other household members), need to be further investigated.

Another important aspect that has not been taken into account is the impact of user characteristics. Previous research has however shown that quality of experience of technologies is the result of an interaction between contextual variables and user characteristics [14]. In a previous step in this research project, three types of household cooks have been defined based on socio-demographic, attitudes towards cooking and cooking behavior (see Sect. 3.4). In a further analysis, we will differentiate results based on these user profiles.