Evaluation of Ambient Light Display for High Information Capacity

Hua Ma

(Suzhou Art and Design Technology Institute,Suzhou 215000, China)

Abstract: Ambient light display presents peripheral information unobtrusively, and it can be context aware and aesthetically enhance specific environment. However, since the abstract characteristic of light, people should attentively keep a good balance between unobtrusiveness and effectiveness, when designing an ambient light display. Especially, in the condition of high information capacity, ambient light display need a mechanism to portray information effectively. In this paper, a framework with an overlay mechanism is evaluated for high information capacity. Based on the framework, two ambient light displays are designed to support the evaluation work. The evaluation questions are related with three aspects including information understanding, transition and encoding, special output channel. From the perception level, interpretation level and experience level, the process of evaluation is implemented. The results indicate that the framework with the overlay mechanism can effectively support varied and specific communicating information. The transitions and information encoding benefit from the proper mapping of states and notification levels in the framework.

Key words: evaluation; ambient display; light; information capacity

Ambient display presents peripheral information in a variety of forms. Intelligent lighting, as a kind of communication medium, has been proved to be appropriate in providing information[1]. Innovative ambient light displays can become context aware and aesthetically enhance specific environment. Ambient lighting conveys information with minimum interference to primary work[2]. So, ambient light displays have been used in different contexts, such as house live, office area and social communications.

However, since the abstract characteristic of light, there are challenges and problems in designing ambient light displays, such as how to keep a balance between unobtrusiveness and effectiveness[2], how to increase the capacity of information visualization, and how to communicate detailed information.

In this paper, our research focus on high information capacity in ambient light display. Information capacity is the number of discrete information sources that a system can represent[1,3]. We describe information capacity with numbers of information dimensions. In our research, two points are concerned, conveying more information without overloading users’ sense and communicating information comprehensibly.

Based on proxemics interactions, we explore a framework with an overlay mechanism to present multiple dimensions of information. Moreover, a special output channel is tried to portray specific information. We believe that it can support ambient light displays in increasing information capacity further.

In order to evaluate the concept, two ambient light displays, “iKnow” and “tLapse”, are created to support the evaluation work. The context is a public space, a public, open and multifunctional room.Qualitative research methods are used to evaluate the effectiveness of communicating multiple dimension information and specific information to users.

1 Related Work

Many previous research work explored different parameters of light (color, brightness, saturation, pattern and frequency of changing) to encode raw data and design display event aesthetically.

“Ambient Orb”[4]is a “ball” with glowing different colors to display real time information, like traffic congestion, weather, stock market trends, and more. It can visualize one piece of information source simultaneously. For example, “Stock Orb”[5]shows stock mark trends as glowing color more green or red to indicate market movement up or down, or yellow when the market is unchanged.

“AmbiPower”[2]try to display the current power consumption with pulses in Philips LivingColors, the household’s remaining power quota with color (green to red depends on the amount of used power). Although, in “AmbiPower”, varying color has proved to be an appropriate way to communicate information unobtrusively, pulsing creates a certain degree of annoying.

Several research explored frameworks of ambient display for interaction phases depending on people’s proximity[6-8]. In these frameworks, transitions create effects on a display[9]from implicit interaction to explicit interaction between different interaction phases, so that users can make requests and anticipate actions[8].

Streitz et al.[6]distinguish three interaction zones, including ambient zone, notification zone and interaction zone. They design different light patterns as transitions from stand-by pattern, notification pattern to represent information from ambient zone to notification zone. They used the mobile device “ViewPort” as a borrowed display to offer services of context awareness, create and visualize detailed and personal information.

“Range” of Ju et al.[8], paid more attention to transitions between explicit and implicit interaction. Design transitions between four zones, public, social, personal and intimate, with implicit interaction techniques of user reflection, system demonstration and override.

Matthews et al.[9], describe peripheral displays as three characteristics, notification, transitions and abstraction. Define notification as five levels: ignore, change blind, make aware, interrupt and demand attention.

2 Evaluation Context and Two Systems

The context of our evaluation work is a public space. “Breakout 404”, is a public, open and multifunctional room for relaxation, conversation in the working space. The frequency of using is high. We try to improve the efficiency of room using. About the room using, people have three aspects of demands. Firstly, it is about the states of room. The door is always closed. So, people need to know occupied or not without opening the door. Secondly, some relevant suggestion cues are needed to help people to make a decision, use the room or not, wait for it or not. It depends on whether it will be free or not. Is there a reservation schedule for the room? How soon will the reservation begin? Lastly, there are demands for people who are in using the room. Time cue can help users manage their behaviors in the room. Also, situations outside of the room may affect people in the room in a certain degree. For example, the frequency of peoples’ asking for the room may affect people’s activity.

When people use “Breakout 404”, they need to know much information. So, we design two innovative lighting systems to meet people’s demands. “iKnow”, outside of the room (Fig.1), can convey information about the states of the room and provide suggestion cues.

Fig.1 “iKnow”—the system outside of the room

“tLapse”, in the room (Fig.2), is designed to show time cues and intensity of room demand. And people can set timeby an application on iPad.

Fig.2 “tLapse”—the system in the room

Fig.4 Assignment of information (for “iKnow”)

From ambient zone, notification zone to interactive zone (Fig.3, three interaction zones of Streitz et al.[6]), people’s attention level and openness for information increase gradually. Closer distances lead to increasing expectations of interpersonal engagement and intimacy[10].

Based on three interaction zones, we explore an overlay mechanism to convey multiple dimensions of information. Different level of information capacity is assigned for each zone (see Figs.4-5).With users’ expectations, multiple dimensions of information are presented. Information overlays the one that has been conveyed in the previous zone that user has passed through. For example, people keep close to the display, when they are in front of the display,they have known from the display that the room is occupied or vacant. The dimension of information, “If the room is occupied or vacant?”, will be potentially and intuitively kept in notification zone and interaction zone, even when it isn’t presented directly in these zones.

Fig.3 Three interaction zones of Streitz et al.

For “iKnow” (Fig.4), we design eight states from ambient zone, notification zone to interactive zone. In the condition of vacant, states ③ and ⑤ stand for that the reservation time is longer and shorter than 30 min separately. Similarly, in the condition of occupied, states ④ and ⑥ represent that the rest time of the activity in the room is longer and shorter than 30 min respectively. State ⑦ conveys the time to the next reservation and state ⑧ indicates the rest time of room occupied.

For “tLapse”(Fig.5), in the room, we assign information depending on people’s behaviors, such as if they are in the room or not (room occupied or not) and whether they set time for themselves or not.

Three states are designed. Once people enter the room, “tLapse” shows state ①, which means people have been detected in the room. If people set time for themselves, system runs state ③.Time cue helps them manage their activities in the room. If people don’t set time, state ② will function. It presents the intensity of room

demand. The times people outside of the room interact with “iKnow” show the intensity of room demand.

Fig.5 Information assignment (for “tLapse”)

Fig.6 indicates how “iKnow” provides information to users in different interaction zones and different conditions. And it also shows how “tLapse” conveys time cues and intensity of room demand.

Fig.6 Information structure of “iKnow” and “tLapse”

3 Evaluation of “iKnow” and “tLapse”

Based on the framework with the overlay mechanism, “iKnow” and “tLapse” are designed to evaluate that if it is possible to increase information capacity for ambient light display and explore how to keep unobtrusiveness and effectiveness when presents diverse and detailed information. These are our evaluation goals.

3.1 Questions for evaluation

According to the evaluation goals, we set up the evaluation on three aspects of research questions bellowed.

Information understanding: Did the information correspond to people’s expectations without confusing them and support them to make decisions? Did people understand the information?

Transitions and encoding: Did “iKnow” and “tLapse” arrange proper notification levels for the system states?Were the transitions unobtrusive and effective?

Special output channel: As a special output channel,can the way of comparing top and bottom part of a lamp express specific time?

Detailed questions were designed from the perception level, interpretation level and experience level, in Tab.1.

Tab.1 Detailed questions for evaluations

3.2 Evaluation setup

Subjects:15 participants were recruited.

Material: Two videos for general description of “iKnow” and “tLapse”. One video of different system settings and a set of open questions (7 question tables) for walkthrough evaluation. Two groups of pictures for time recognition evaluation of “iKnow”.Three groups of pictures for time recognition evaluation of “tLapse”.

3.3 Implementation

We use the qualitative research methods, interviews, questionnaire and observations.

Introduction: Two videos were displayed to do general descriptions about “iKnow” and “tLapse”, before the true test.

For “iKnow”: After an introduction, the participants should answer the questions about the information that they need from “iKnow” in different room conditions. Then, they became to experience “iKnow” by video and in the real environment. Seven questions tables about recognition, notification and information interpretation were asked separately when the subjects explored “iKnow” in different conditions. During exploration, they were asked to estimate time from 4 pictures without order and 3 scenes in video with order for time recognition, in red and green color separately. After exploration, to evaluate helpfulness of “iKnow”, the participants should complete two tasks of making decisions of using the room or not and waiting for the room or not.

For “tLapse”: The subjects were asked to finish the first task in the room. They need to set time of 5 min by iPad. Then “tLapse” began its transitions. During the 5 min (“tLapse” was in transition), they should finish the second task about time recognition, estimating time from 15 pictures, which in three colors (red, green and blue). Every color has 5 pictures. After the second task, the subjects had two questions to answer, which were about distraction of “tLapse”. When time was up, the transitions of “tLapse” finished. After a while, the transitions for room demand appeared. At this moment, the subjects were asked if they could understand the meaning of the transitions.

3.4 Data analysis

We collected data from 15 participants, including oral and literal answers and the results of their tasks. We use a quantitative method to analyze the results.

About notification levels and transitions: The findings from recognition and notification indicate that the subjects can catch and notice the transitions of “iKnow”. They can notice that the brightness increasing appeared because they were in front of the door and they would focus on “iKnow”, show their curiosities when LED flashing. By increasing brightness and flashing continuously, “iKnow” attracts people’s attention gradually from inattention to focused attention[7]. Two participants, however, also showed their curiosities when “iKnow” increased brightness only (without LED flash), especially in the condition of vacant. It points out that in the notification zone, although without focused attention, there may be opportunities for people to obtain much more information. Furthermore, for the two participants, when vacant, they need a little lower notification level than when occupied. It shows us cues to explore that is it possible to arrange different notification levels in the same zone when the conditions are different (vacant and occupied). The results from distraction show that “tLapse” didn’t distract them from their main task in the room. The smooth transition of “tLapse” can help people to track the cues of time and room demand when they are not focusing on their main tasks.

Information interpretation: The outcomes related to information interpretation illustrate that, for “iKnow”, the subjects can well understand that flash is not only attracting their attention but also indicating that the room will be occupied or vacant and will tell them how soon. So, after they touch it, even “iKnow” changes the data mapping, they can still understand what it means. (This comes from subjects’ real experience when they were in front of the room, but when they only saw movie, it seemed a little difficult.) As a result, the participants can understand the information conveyed by “iKnow” without confused. However, after knowing about the whole system, some subjects showed their interests in obtaining much more information from “iKnow”. There’s one participant would like to know more in ambient zone. It means that not only the information of occupied or vacant but also if it will be vacant or occupied soon as well. For “tLapse”, participants can recognize when it’s presenting time cue and when it’s conveying room demand from outside, based on time setting or not, even “tLapse” uses the same transition. But there’s one participant who worried about that she will forget if she has set time or not. So, she hoped there would be differences when conveying different information. It could be explored in the future that people’s affordability of the differences between transitions.

Helpfulness: The results from the two tasks for helpfulness indicate that all the subjects can make decisions of using the room or not and waiting for the room or not, depend on the information visualization. Although subjects believed that the information portrayed by the systems are helpful,several participants understood “soon” with different standards. Some considered “soon” as 5 min, some believed it as 10 min and some thought it as 60 min. These standards show that people have different understandings of “soon”. So, define less than 30 min as “soon” is useful for some subjects but not for everyone.

Recognition of specific time: We calculate error ratios by formula ofR(E)=ABS(V(R)-V(A))/30.V(R) means the real value of time in the picture.V(A) means the average value of all the subjects estimated from the picture.R(E) means the error ratio of every picture. The analysis of time estimates is compared in red and in green (Figs.7-8). The outcomes indicate that errors in red are smaller than that in green. So, it’s obvious that it can be a way to show time, and red color is much more readable than green color. People can estimate time from it, especially recognize the conditions of more than 15 min and less than 5 min left. The results show that as an output channel the comparing way is useful at the beginning and the ending of timing.

Fig.7 Time estimate for “iKnow” in red

Fig.8 Time estimate for “iKnow” in green

4 Conclusions and Future Work

To evaluate ambient light display for high information capacity, we design two systems on our concept of framework with an overlay mechanism. During the evaluation, people show their curiosity about the two systems and they feel it novel and helpful to obtain useful and diverse information from a light.

4.1 Conclusion of evaluation

The exploration and study results validate that the framework with the overlay mechanism can support ambient light display communicating varied and specific information. At the same time, the overlay mechanism makes it possible to add data sources, increase output channels and change data mapping to convey multiple dimensions of information without confusing users.

Transitions design and information encoding with light benefit from the proper mapping of states and notification levels. People can catch and notice the transitions of system based on their own behaviors and room’s different conditions. The transitions succeed in attracting user’s attention in different levels at different moments, in order to portray diverse and detailed information.

The specific output channel is a kind of way to indicate specific time. It is much more useful at the beginning and the ending of timing.

4.2 Future work

People’s wishes for much more information motivate us to do further exploration about high information capacity for ambient light display. We would try to concern more about contexts when arrange notification levels and design transitions in the framework.Besides, in the future, it is important to explore people’s affordability about different transitions when interpreting more information in every proxemics interaction zone. About the touching area on the light, it is interesting for us to investigate how to imply people to touch a light without informing them and to explore the possibility to present specific information without people’s touching, after succeed in attracting their focused attention. Lastly, for the specific output channel to indicate specific time, we would like to explore the readability in different colors.

Acknowledgments

Particularly grateful to professor Berry Eggen and professor Harm van Essen, who guide me during my period as a guest researcher in Tu/e. They strongly supported and helped me about my research work. Especially in the program of evaluation, they gave me so much important advice.

Thanks to my colleagues in yellow space of ID. They gave me a lot of suggestions in our discussions and help me much more during prototypes making.