PostBits: using contextual locations for embedding cloud information in the home
- First Online:
- Received:
- Accepted:
- 127 Downloads
Abstract
Placing information at specific locations in the home provides rich and intuitive ways for people to cope with information, as they leverage semantics of the locations within the home. However, there is no deeper investigation yet on how users would embed digital cloud-based information into various locations in their homes, partly because previous systems were not robust enough to be deployed in real settings for an extended period of time. To this end, we have developed PostBits, a system of display blocks that integrate cloud information with contextually rich physical space. PostBits were designed for long battery life, robust communication and simple interactions, to enable a field deployment. A field study was conducted with 6 families, each using the system in their home for 1 week. We have identified patterns and strategies of how users embed cloud information at contextual locations in the home, and reflect on future design opportunities.
Keywords
Pervasive displays Ubiquitous computing Smart home User interfaces1 Introduction
Traditional media such as handwritten notes and paper calendars are still used extensively in domestic settings as effective means of communication [4, 9]. Although this inexpensive and ubiquitous medium may seem primitive, the ability of contextualizing the information adds significant amount of richness. For example, a post-it note left on a wallet would provide a just-in-time reminder to pick something up before leaving to work. In fact, previous research has shown that information such as reminders/alerts, schedules and notices is created and understood by home inhabitants as a function of contextual locations within the home [9].
On the other hand, digital information on the cloud brings other significant advantages such as searchability and ability to update itself. However, accessing information on the cloud through a single block of screen on a personal device (e.g. mobile phone, smartwatch or personal computer) misses the important cues provided by the contextual locations. There is a large body of small and everywhere displays proposed in the literature (e.g. [3, 7, 8, 12, 18, 21, 22, 25]. However, to our knowledge, there is limited understanding of how such systems would be used in a home setting for an extended period of time. Such real-world understanding is critical to guide the design of future situated display interfaces.
First, we share findings from a field study conducted with six families to understand how non-expert users would use, manage, deploy and redeploy PostBits in their homes. We visited each family three times during a week to conduct semi-structured interviews and observations. Our findings suggest a set of unique patterns and strategies that emerged when the participants used PostBits to integrate information on the cloud to their homes through contextual placement. We highlight how these emerging usage patterns differentiate PostBits from traditional media. For example, one family had multiple PostBits in the living room connected to Twitter feeds of two political parties which kept them aware of the ongoing general election. We describe implications of PostBits system in a home setting and explore further design opportunities.
Second, we share the technical details of the PostBits. Elliot et al. [10] have indicated the technical challenges of developing a prototype that is ready to be deployed in a home setting. We developed customized hardware and made robust PostBits prototypes that are energy efficient and reliable enough to be left with end-users at home. Also we developed a scalable communication hierarchy and content management system to embed both user-generated and publicly available information with PostBits. Our current custom-made PostBits are energy efficient to operate for more than a week without recharging. This technical development made the domestic evaluation of PostBits possible.
2 Related work
2.1 Contextuality of information at home
Home is characterized by a variety of user groups that differ in age and general abilities [5, 19]. Previous examinations of households revealed the importance of the routine work of communication [4]. Furthermore, locations of information used in communication are often determined by the daily routines of the inhabitants [9]. Also certain “typical places” within home are designated for specific family members that provide information on the social organization of communication within the household [4]. It is important to understand the routines and behaviours of the family members in order to place contextually sensitive information [6]. Elliot et al. [9] used contextual semi-structured interviews to reveal that the main types of communication information at home shared using paper-based and electronic media included “reminders and alerts”, “schedules”, “notices”, “visual displays” and “resource coordination”. Chetty et al. [2] examined the relationship between home networking and the house itself—how technologies interact with the house infrastructure and how it affects the householders.
These studies show the importance of the contextual location of the message displayed and how it attaches meta-information about action, activity, time, ownership and awareness. These studies are limited to exploring how “traditional” media are used within a home context. With PostBits, we aim to investigate how “small and everywhere digital displays” perform in conjunction with existing physical practices as well as the unique advantages they offer in comparison with traditional media.
2.2 Small and everywhere displays
Early on, research has explored and verified the efficiency of increasing communication and awareness in collaborative workplaces through ambient and contextual information interfaces. Dourish and Bly presented a system to increase awareness through Portholes, a regularly updating image bulletin [8]. Fitton and Cheverst showed how an office door display can enhance the awareness and communication utilizing the location of display [11]. Notification Collage is another example of how a secondary display monitor and strategically located public displays enhanced office collaboration and communication [13]. However, these interfaces lacked the mobility to change their location according to the user needs.
In home environments, one approach to leverage on contextual nature of information is to embed existing domestic objects with information. Hazlewood et al. [14] showed how domestic lights can act as an ambient communication medium. Casablanca is another example of embedding household object with information and extending it to other mediums such as sound [15]. Furthermore, Mynatt et al. also demonstrated the utility of sounds in ambient communication context [18]. Due to the limitation of expressivity of the medium, types of information displayed through these interfaces were few and abstract.
Recent efforts in academia on small and everywhere displays
Publication and year | Name | Display type | Focus | Development | Evaluation type | Ubiquity |
|---|---|---|---|---|---|---|
Elliot et al. [10] | Sticky spots | Existing displays in homes | Home | Conceptual design | None | |
Fitton et al. [11] | Hermes office door display | Custom-made displays | Office | Special set-up | Field study | No |
Saslis-Lagoudakis [25] | Hermes@Homes | Custom-made displays | Home | Special set-up | Field study | No |
Greenberg [13] | Notification collage | Existing displays | Office | Conceptual design | Laboratory study | No |
Kalanithi [16] | Connectibles | Custom-made widgets | Home | Laboratory set-up | Laboratory study | Yes |
Ziola et al. [31] | Desk Jockey | Projection | Office | Special set-up | Field study | No |
Olberding et al. [21] | CloudDrops | Custom-made displays | Home | Early prototype | Pilot | Yes |
3 PostBits
3.1 System design and interactions
3.1.1 System overview
System overview and architecture
3.1.2 Interactions and feedback
Interacting with PostBits by a shake b swipe and c switch ON/OFF. Shaking the PostBit will d reset a PostBit, Swiping will e refresh the display and f if the PostBit is out of range it displays the message
3.1.3 Content management
Accessing PostBits through the web UI and making content updates
Different types of user input and information feed currently supported by PostBits
3.2 Implementation
3.2.1 PostBits
Internal electronics of PostBits
3.2.2 Server back end
Server side back-end application was implemented using Java and deployed in a production-level application server which is hosted on an Amazon EC2 virtual machine. At the core of the back-end application, an image processing module was implemented to transform contents into the format supported by the E-Ink display. Text formatting and image re-scaling operations were implemented directly using native Java2D functions. Custom-implemented image binarizing and image dithering algorithms were used to convert colour input images to PostBit-supported binary format. A MySQL database server that runs on the same EC2 instance was used as the data persistence mechanism.
3.2.3 Content management front-end
Web UI was designed to provide an easy interface to manage the contents in PostBits. We implemented the front-end user interface using JQuery mobile Web framework, because of its compatibility with all major desktop and mobile browser applications.
3.2.4 Communication
We used ZigBee low-power short-range wireless communication mechanism between PostBits and the intermediate router. Commercially available programmable XBee to IP gateway device is used as the intermediate processing unit. A Python program was implemented to fetch image data via REST API from the server as per the incoming requests from PostBits. Since low-power XBee modules are not capable of transmitting a complete image of nearly 6 KB at once, we used the memory of the gateway device as the intermediate cache to store incoming image data from the server. The stored image data were then sent to the PostBits chunk by chunk, 64 bytes each. In addition to the image requests and responses, shake and touch events were also being sent to the remote server via REST API through the gateway. HTTP over TCP was selected as the protocol for connecting the remote server. This gateway is connected to the Internet through users domestic Wi-fi or Ethernet network.
4 Field study
4.1 Participants
Summary of study participants
ID | Family composition | House structure |
|---|---|---|
F1 | Husband—working professional (34 years), wife—homemaker (37 years), one child | 2 bedrooms |
F2 | Husband—working professional (34 years), wife—homemaker (32 years) | 2 bedrooms |
F3 | Husband—working professional (32 years), wife—homemaker (26 years) | 2 bedrooms |
F4 | Husband—working professional (34 years), wife—homemaker (34 years), one child domestic helper (43 years) | 3 bedrooms |
F5 | Husband—student (28 years), wife—homemaker (26 years) | 1 bedroom studio |
F6 | Husband—working professional (33 years), wife—homemaker (36 years) | 2 bedrooms |
4.2 Procedure
At the start of the study, each family was given a demonstration of how the PostBits work and were handed a set of 5 PostBits with the option of asking for more if they needed. We conducted three semi-structured interviews, each lasting 30 min, with family members from each household, one at the beginning, one after 4 days into the experiment and the last one at the end of the week. Semi-structured interviews included a tour of the home, in which participants showcased how they had placed the PostBits and who was using them. For pragmatic reasons, we did not interview children/teenagers below 18 years.
Initial interview explored the structure of the participant’s family and existing communication means they used at home. Core questions of the intermediate and final interviews revolved around how the participants used the PostBits, the number of PostBits used, types of content frequently shared, who used them often and whom the messages were intended for. We also obtained their opinion on why they found certain features of PostBits favourable, what did not work for them and their suggestions for what they would like to see. In addition, we captured photographs of the PostBits as placed in various locations at home. Moreover, we logged the contents of the PostBits via the server to get a deeper understanding on the interview data.
4.3 Data analysis
The interviewers took detailed handwritten notes during interviews and home walk-throughs. We open coded the data and qualitatively analysed the observations of user behaviour and user reports from interviews [1]. During the study, F1 faced Internet connectivity issues at home, thereby making server logs inconsistent. As such, we have not analysed the server logs from F1 to avoid any discrepancies. However, we have analysed the interview data of F1.
5 Findings and discussion
From the semi-structured interviews and server logs, we observed several usage patterns and strategies in terms of how participants deployed and used PostBits. We discuss findings with reference to parameters revealed in existing literature such as integration within architectural space, sense of awareness, ownership and urgency. In addition, we identified a set of emerging usage patterns with PostBits including spatially directed remote postings, active in situ communication and spatially filtered information feeds. In the following paragraphs, we summarize these findings and offer comparison to traditional media.
5.1 Existing information systems at home
During the first semi-structured interview, we explored the existing information systems participants are already using at home. We analysed them along two aspects: type of content shared and the tools used to share the content. The participants shared all the content types revealed by Elliot et al. [9], namely reminders/alerts, schedules, visual displays, notices and resource coordination. Out of these, reminders/alerts were the most dominant type. These messages were conveyed using a variety of media like sticky notes, handwritten notes, text messaging through phones and emails. For example, F2 reported that they use “to-do lists” on the fridge doors in the form of shopping lists or notes of what is inside the fridge and little pin-up notes at study desk mostly with reminders about upcoming meetings. F3 reported to use smartphone applications for resource coordination for activities like shopping and other household chores. However, most of the families shared that for urgent communication and messages, they often resorted to phone calls to the relevant person directly.
5.2 PostBits in the home environment
5.2.1 Integration with architectural space
Assigning specific content types to PostBits based on location
Duration of display of information feeds and user input
5.2.2 Number of PostBits
Throughout the study, none of the participants asked for extra PostBits than the 5 given, even though they were given an option to ask for more. Some of the users attributed this to the size of their apartment, while others attributed it to the fact that there were only 3–4 locations in the house where they thought placing messages made sense. For urgent messages, they continued to use other media like smartphones. This could also be due to the relatively small size of the households, usually 2 inhabitants with the exception of F4 (4 members). Further, F1 and F3 used PostBits as complementary device, along with traditional media. While F1 used paper notes for lists and PostBits for information feeds, users from F3 used both sticky notes and PostBits for reminders depending on what was more convenient at that point in time. The shared use of PostBits and sticky notes for certain message types may also have impacted the number of PostBits used as families did not use PostBits for every message. If the users found a pen and paper more accessible than their mobile phone or laptop, they resorted to writing a reminder on a paper. As PostBits are ubiquitous, the inhabitants were able to reconfigure domestic spaces and the PostBits to meet the demands (e.g. [24]) such as moving them from the kitchen platform to the fridge depending on the content. For example, F6 shared that, “I leave a PostBit on the fridge that reminds my wife to take the food out from the freezer. After she removes the food, she sometimes puts it on the kitchen platform with a different message” (Fig. 9b).
5.2.3 Type of content
Frequency of content-change by users
5.2.4 Sense of awareness
PostBits have the ability to retrieve assigned cloud information automatically and update itself with relevant content. The fact that they can assign these information feeds into specific locations differentiated the PostBits from other traditional media. For example, F4 had multiple PostBits in the living room, assigned with different Twitter handles. In fact, F4 mentioned that it was convenient to be able to see tweets from the two main parties contesting in the election (study was conducted during the general election) near the television. Also, we observed that some people moved the weather feature to their private space to know the weather of places where their loved ones were travelling or living. One of the users from F2 shared, “I really liked the weather feature. My wife is mostly in Tokyo and I would like to see the weather of the place or before her flight. If I am alone, I may even keep this PostBit with weather on my study desk” (Fig. 10a). Moreover, one of the participants from F6 shared that it is important for her to have a sense of the air quality index for the day. In fact, they had visitors during the study and the host used one PostBit in the living room showing local weather conditions so that visitor could better plan sightseeing. This also revealed how daily life practices were build around a working data connection [17].
5.2.5 Sense of ownership
We observed specific patterns in how content varied across “public” (accessible to all members of the family and no dominant user) and “private” (dominant user). Information that determined the planning of the personal schedule, was placed in more private spaces as compared to information for resource coordination that was always placed in more “public” spaces such as kitchen and the living room. This was logical given that resource coordination was among different members, while personal schedules seemed more relevant to specific individuals. For example, one of the users from F2 pointed to a PostBit on his study desk and shared, “This is my PostBit. I work here very often. Sometimes, I keep a reminder on the PostBit and before I leave for work, I just look at it to know when I have a meeting etc. I do this with sticky notes too. I am used to checking reminders here” (Fig. 10b). It was interesting to see how users referred to the PostBit in their private space as “my PostBit”, one that only they would change. This is in contrast to the PostBits that were placed in the living room or kitchen that had shared ownership. For Twitter feeds, we observed that more general tweets such as news channels were placed in public spots (Fig. 10a), while personal favourites like sportsman or scientists were placed at private areas like study desk. If the message was meant for a specific member in the family, then the PostBits were placed in the private space of that person.
5.2.6 Sense of urgency
Shared use of PostBits between family members and reconfiguring the PostBits based on location
6 Emerging usage patterns of PostBits
We observed three emerging interaction scenarios where users leveraged on unique features of PostBits (e.g. remotely updating the contents, linking to an information feed), which are not available in traditional media such as sticky notes.
6.1 Spatially directed remote postings
Prime sites for placement of PostBits—a dining area/living room, b study, c bedroom and d kitchen
6.2 Active in situ communication
PostBits enabled Active in situ communication: sharing and updating information on a display situated at home. This information helps not only the user who posted it but also the other members of the family. For example, participant from F2 shared that he always left a PostBit at his study room. According to him, “…over the course of the day when meetings change, I just change the text on my PostBit immediately from the office instead of having to write a sticky note (after reaching home)”. His wife shared, “When I am not sure whether my husband is coming for dinner or why he is late, I sometimes check his PostBits”. This was particularly novel as the content was used to infer about the user’s presence, absence and predict their arrival. Since she knew that the content was more likely to be updated, there was also a certain element of trust that she as a user placed on this content. Another participant from F5 shared that, “With lists, I found it really convenient to delete things off the shopping list once I purchased it”. This particular PostBit was attached on the kitchen refrigerator and helped the decision-making process of the family member at location (wife) to decide whether she needs to go out shopping and if so what she needs to buy. In this regard, updating content on the PostBits placed at a designated “public” location enabled situation awareness.
6.3 Spatially filtered information feeds
The fact that the PostBits connected to an information feed (weather, tweets, news) can be attached to a specific location has been used as a filtered information feed to that location. We identified this as spatially filtered information feeds. For example, one of the users from F6 allocated two PostBits for Twitter feed. While one PostBit carried the feed of his mentor and was placed on his study desk, the other was from a sportsman he is a fan of and was placed in the living room. He said, “I follow my mentor ‘X’s’ tweets closely. He inspires me. It is really cool to have this auto-update feature, now I can always see the latest tweets”. He liked to see his mentor’s tweets while he studied (context), but not the sportsman’s tweets. However, in the living room, he preferred to get away from work and just focus on the entertainment. Another user from F4 used two PostBits to monitor the ongoing election updates from the two parties involved. He chose to place both PostBits in the living room where they also had the television that displayed news related to the election. With these usage patterns, we can infer that there is a filtering of information feeds using location (context) as the key.
6.3.1 Novelty effects
Our study highlighted how PostBits are used in home settings for a period of 1 week to communicate a variety of content. We observed novel patterns and strategies during this period. In addition, with an aim to understand how this usage changed over longer periods, we conducted an extended study with F6 for 3 weeks. We found that with increased time, the users got more familiar with the device and were able to assign content with ease through the Web interface. In addition, they reported to be able to use PostBits as a learning tool with more ease. We observed that the participants assigned learning content to each PostBit and placed them in the same “prime” sites (kitchen, bedroom, living room). One of the interesting patterns that emerged was the fact that some PostBits were switched OFF after they placed important information such as an image that they did not want to be changed. We plan to run an extended study with more families to understand how current usage patterns would change over time and if newer strategies would emerge.
6.3.2 Direct manipulation of content
One user from F1 shared, “When making a quick list of things to buy or leaving notes for someone, it is intuitive to reach for pen and paper. Or anything to write with. Having a stylus will be really a good addition”. This suggests that users tended to use the medium that was most convenient at a given point in time, especially for sharing quick messages. This was affirmed by other users who also shared that they sometimes find it “handy” or more intuitive to jot down a quick message on a piece of paper. These observations emphasize the need for incorporating a mechanism for direct input into the PostBits such as being able to write on the PostBits screen itself.
6.3.3 PostBits in other contexts
In order to observe how PostBits are used in other contexts such as shared workspaces, we have initiated a preliminary study at a research laboratory setting consisting of 3 individuals. We observed that PostBits were used to communicate ideas between people such as leaving notes/reminders for others and updating progress on joint tasks. The users reported that they found this quite different from sticky notes only when there was a change to be updated, and they were away from their desk. However, the users also felt the need to have a more interactive/scroll function on the screen as they felt that the screen space was too small to leave certain long messages. The PostBits were not moved around much during the study as the 3 users divided the PostBits among themselves. Since the weather in the area was unpleasant during the study, the users also reported to have found the auto-update of weather function useful. The users in fact felt that the low notification level was good as it did not interfere with their work, and they could glance at the message when they felt the need to. However, they also shared that if the notification can be controlled by the user, it would be useful in some contexts. In addition they felt that it would be beneficial to know whether the intended recipient had seen the message and acted on it.
7 Design implications
Further design opportunities of PostBits-like systems can be discussed with a taxonomy of ambient information systems [23]. This includes four design dimensions:
7.1 Notification level
The degree to which system alerts is meant to interrupt a user. With PostBits, we kept the notification level to a minimum where the user does not get any indication about content changes. This meant that there were no beeps or vibrations to notify users of any update. During our interviews with users from home and work settings, only a couple of users indicated the preference for obvious notification. Therefore, when designing similar systems, the absence of a notification mechanism can be argued as both a limitation and a feature, as some users preferred PostBits as a non-disturbing communication method. However, in future designs, implementing an optional notification mechanism would be worth considering.
7.2 Information capacity
Number of discrete information elements the system can display. User feedback and study observations indicate that cloud-based information feeds were largely placed in contextual locations and potently utilized in the domestic environment. Accordingly, providing an open platform to integrate a wide range of cloud-based feeds rather than using a fixed set of information feeds will further enhance this feature.
7.3 Representational fidelity
How the information is encoded into the representation medium of the display. Compared to other features, image content was less utilized by the users. This may be due to the relatively small and monochrome display of the PostBits. However, supporting rich image content with a larger, colour display would result in trade-off in terms of power consumption and portability. Even though a conclusive point cannot be made in this regard, the ability to display rich image content in such systems does not appear as a key requirement.
7.4 Aesthetic emphasis
How visually pleasing an object is when placed in the environment. All the users found PostBits to be aesthetically pleasing and resembling sticky notes. We believe that this was one of the reasons why they found it easy to integrate PostBits in their daily lives and architectural space. PostBits were placed on various locations at home, and one PostBit was broken as it had dropped on the floor. As such, we identified the need to create good attaching/supporting mechanisms. However, this needs to be done without making the design cumbersome. We found this as a key design consideration that would facilitate fluid integration of the system with the physical context of the home.
8 Conclusion
In this paper, we present a pervasive platform, PostBits, to investigate how users would integrate digital information on the cloud onto contextual physical locations in their homes. Implementation of the PostBits was robust and energy efficient to be operated for an entire week without recharging. The PostBits could be assigned with user inputs (text, list and images) that need to be manually updated or information feeds (weather, twitter and news) that get auto-updated. We gave PostBits to 6 families where each family used the system for a period of 1 week. We found that the usage of PostBits were similar to traditional media as users placed contextually relevant information on them. The study also revealed unique usage patterns and advantages of PostBits in comparison with traditional media, namely spatially directed remote posting, active in situ communications and spatially filtered information feeds. We also conducted a long-term study with one family and explored the use of PostBits across office and workspace contexts through a preliminary study. Our findings and observations motivate us to continue user studies over longer periods of time in home and office settings. We believe PostBits-like system would provide an intuitive way to connect digital information on the cloud with the physical information in our living spaces.