Usability Study of Indoor Mobile Navigation System in Commercial Facilities Yutaka Inoue, Takeshi Ikeda, Kiyoshi Yamamoto, Tomohisa Yamashita, Akio Sashima and Koichi Kurumatani National Institute of Advanced Industrial Science and Technology (AIST) / CREST, JST 2-41-6 Aomi, Koto, Tokyo 135-0064, Japan yutaka.inoue @aist.go.jp ABSTRACT For a usability study of indoor navigation services, we con- In this paper, we describe a field experiment of an autono- ducted a field experiment in large-scale commercial facili- mous mobile system to navigate users and to estimate the ties to investigate the usability of the indoor navigation ser- self-position indoors, where it is impossible to receive a vice on a mobile terminal. Our indoor navigation system GPS signal. Many users utilize positional information sys- might be developed so that it can guide a user around com- tems of GPS. However, because positional information mercial facilities. Moreover, the system must be evaluated computed using GPS signal is available only in the loca- by typical users. tions with satellite signal reception, it is difficult to use GPS for positioning in indoor environments. Therefore, we have We describe the usability of an indoor navigation system on developed an indoor navigation system to provide ubiqui- a smart phone to use position information in an indoor envi- tous information service like that of a portable navigation ronment, and results of the experiment. Furthermore, we system that can be used inside commercial and office build- discuss effective indoor navigation functions and the inter- ings. The navigation system can display user’s position that face considering users’ opinions obtained through the ex- is estimated by beacon signal using license-free radios. We periment. carried out a field experiments in large-scale commercial FEATURES OF THE INDOOR NAVIGATION SYSYTEM facilities to evaluate the usability of the navigation interface and availability of the indoor navigation service on a smart We have developed an indoor navigation system consider- phone. As a result, we found that the users would like to ing usability. The indoor navigation system service has the have some functions of indoor navigation systems. following features. ● Showing the current position of the user in a floor map. Author Keywords ● Changing the floor map according to the user’s position. Positioning, Position estimation, Indoor navigation, Smart- ● Showing routes from a current position to a destination. phone, Cell phone, Wireless beacon Figure 1 shows a user’s current position expressed using a circle in the center of the display screen. As the user moves ACM Classification Keywords from left to right on a floor, the user’s position in the screen C.5.5 [Computer System Implementation]: Services is updated by the navigation system. Moreover, when the user moves to another floor, the system automatically rec- INTRODUCTION ognizes the current floor and changes the user terminal map When considering various services for mobile terminal us- to the floor map. ing systems like our indoor positioning system, we must consider what services the users’ preference, functions of Usually, a navigation route is displayed on a floor map on the service, the operation interface, and the screen view. For the terminal screen as shown in Fig. 2(a). When the user instance, human-navigation services in outdoor environ- reaches the points of navigation events, such as an intersec- ments rarely require notice to users about vertical move- tion on the route or in front of an escalator or elevator, a ments or positions. However, in indoor public spaces, guidance picture like that shown in Fig. 2(b) is displayed on where it is impossible to use GPS, such as commercial fa- the screen along with a beep sound. The screen presents cilities, users might use elevators and escalators to go to the guidance to lead the user to the destination. The guidance is target location. shown using multimedia: natural language, pictures (e.g., icons, arrows), and photographs of the location. Conse- Copyright is held by the authors. quently, the user can reach the destination merely by fol- UbiComp '08 Workshop W2 -- Ubiquitous Systems Evaluation (USE '08) lowing the guidance. Advice by the screen announcement September 21st, 2008 This position paper is not an official publication of UbiComp '08. includes right turn, left turn, floor changing via escalator or elevator, and destination arrival. 1 transmitted from radio beacon devices installed in indoor environments. The radio beacon device is small and light- weight. Therefore, it is possible to attach it in an inspection door on the ceiling. A beacon receiver to be carried with the user is small and cable-less. EXPERIMENT IN COMMERCIAL FACILITIES In this experiment, we intend to collect diverse opinions from ordinary users. The users actually operated the user Figure 1. Example of Positioning Result. terminal of the indoor navigation system in the building and reported widely various opinions about the navigation sys- tem. Experiment Environment We conducted an experiment to investigate users’ demands for the service in a large-scale commercial building. The Please go to the building is known as not only as a shopping center but also diagonally forward left, left and up to by escalator. escalator 【Restroom for babies】 as a sightseeing area in Yokohama City. We selected the building so that the participants in the experiment would be Please go to the 4F Here is baby’s room fourth floor. for changing diapers. menu able to seek and obtain various services in the real world. (a) Map View. (b) Navigation View. (c) Arrival Guide. To guide the users to various destinations, we created a database of locations in the building, e.g., stores, bathrooms, Figure 2. Example of Routing Guide Screen. sightseeing spots, and so on. A smart phone with a touch pen interface (FOMA M1000; NTT docomo Inc.) was se- lected as the target terminal with which a user can input The battery- battery-driven radio beacon devices are many destinations graphically. The interface is highly inter- installed in environment. Beacon devices active; the user merely pushes the intended image on the transmit beacon signals (license- (license-free radios). screen, such as a picture of a store, using the touch pen. Beacon2 Beacon3 The field experiment was conducted at Yokohama Land- Beacon1 mark Plaza building, which is adjacent to the Yokohama Landmark Tower1. As shown in Fig. 4, Yokohama Land- Beacon6 mark Plaza has five floors; each floor in the building has Beacon4 shops, restaurants, hair salons, and so on. The floor area is Beacon5 about 10,000 [m2]; the center part of a floor is the blow-by. Therefore, the user can receive the positioning signal sent User’s device by the radio beacon devices installed on each ceiling on User’s device User’s device other floors. Radio beacon devices were installed in the second floor, the third floor, and the fourth floor: the user can use the naviga- tion system from the second floor to the fourth floor. The radio beacon devices were installed at a ratio of one unit for about 200 [m2] on those floors. It is not necessary to install Bluetooth the radio beacon devices uniformly because the user can connection specify a self-position if the beacon receiver carried by the user can receive one or more beacon signals. The field experiment was carried out for five days during November 2007. Several kinds of participants took part in Beacon signal receiver the experiment. The disaggregated data are of 37 partici- Mobile information device pants: 5 students, 27 homemakers, and 5 elderly people. (Cell phone or smart phone) The participants in the experiment move inside the building, using the smart phone and beacon receiver, as shown in Fig. Figure 3. Architecture of Positioning System. 1 As depicted in Fig. 3, the position is estimated on a smart Mitsubishi Estate Co. Ltd.: The Landmark Tower Yoko- phone with a beacon receiver that receives signal data hama 2-2-1, Minatomirai, Nishi-ku, Yokohama, Japan. 5, and operate the indoor navigation system by referring to countering a cross over points on the displayed route, the user manual. before riding an elevator or before riding on escalators. The participants experienced one or more scenarios among C) A question about whether the user was able to input 11 scenarios prepared in advance. Each scenario assumes a easily a setup of the destination shown by the naviga- situation and purpose of the navigation, e.g., shopping with tion menu. young men and women, sightseeing with the family, and so D) A question about whether the indoor navigation service on. Users chose scenarios that were suitable to their own with the user’s cellular telephone would be useful–– attributes, and moved around in the building according to whether the user would like to use the service. the scenario using the navigation system. The questionnaire results in the experiment are shown in Fig. 6. (a) Navigation screens. 4F Elevator Regarding the questionnaire result (Fig. 6(a)) about guid- Escalator Escalator Escalator ance of the navigation displayed while the user moves to a destination, about 30% of participants responded that the indicated direction can be recognized by comparing the 3F photograph image of the screen and the surroundings. More than the half understood by comparison several times. That is, we understood that 80% or more of the participants were able to reach to the destination without becoming lost; 2F it was possible to select a correct passage by the navigation system. Therefore, it is considered that guidance to a desti- nation using the navigation system indoors was effective. Figure 4. Floor Map of Yokohama Landmark Plaza. (b) Display timing of guides. As shown in Fig. 6(b), from the questionnaire related to the viewing timing of the guidance advice, about 60% of par- ticipants answered that the displayed screen showed timing neither good nor bad. We considered the following as factors of the result. 1) In our system, a guidance picture is displayed at a point that is distant from a fixed distance, immediately before an intersection, near an elevator or an escalator. Therefore, when the guidance picture is shown at a lo- cation with no turning point, it is also true that the us- ers might have difficulty recognizing the point. We are developing a technique for adjusting the guidance points to resolve this problem. 2) As another factor, it is considered that each user has Figure 5. Indoor Navigation Image in Experiment. different timing to watch the smart phone screen; the timing changes from moment to moment. For example, although a user stopped to check the guidance screen, Experimental Results To investigate the usability of the indoor navigation system, in subsequent guidance, the user checked the screen we administered the following questions to solicit opinions for subsequent guidance while walking. In such cases about the experiment. of a walking check, the user felt that the guidance screen was shown later than when they stopped to A) A question about whether the user understood that they check because the user had moved forward several were led in some direction by watching the map and meters by the time they recognized the guidance guidance picture on the screen of the smart phone. screen. Moreover, a fast walker might be shown the screen much later. To solve that problem, we consider B) A question about whether the guidance picture was appending function that modifies the timing for the displayed at the right time; in other words, the user guidance according to the walking speed acquired in was able to watch the picture immediately before en- real time by another sensor. 3 0% 0% 3% 0% 3% 11% 5% 11% 30% 19% 56% 62% Notifyed immediately after watching screen. Always, opportune notice. Notifyed after compairing screen with arround present location. Generally, opportune notice. Notifyed after sevral compairing screen with around present location. Half-and-half opportune notice and ill timed. Notifyed after compairing screen with around present location and walking around. Generally ill timed. Not notify after compairing screen with around present location and walking around. Always ill timed. The other. The other. (a) Guidance screens. (b) Guidance timing. 3% 0% 3% 3% 0% 3% 19% 19% 19% 35% 40% 56% Possible to input data immediately and it is very useful. I would like to use, even when the receiver is required. Possible to input, generally understood how to operation. I would like to use, if the receiver is miniaturized. Although usage is unclear, possible to operate, I would like to use, if the receiver is unnecessary. Although ask staffs for the experiment about usage, data input operate own. Neither necessary nor unnecessary. Impossible to understand, thus all input operation entrust to staffs. I do not want to use it. The other. The other. (c) Navigation interface. (d) Usage of cell phone. Figure 6. Result of Questionnaire in the Experiment. 3) The difference of guidance timing also occurs from operating instructions and used the navigation function (Fig. precision error. The error results from the difference 6(c)). between a user’s position and estimated position in the We consider that anybody can use the input interface easily system. Although the precision of location estimating because the interface of the smart phone design is easy to changes according to number of radio beacon units in- understand. For instance, the destination is chosen by stalled in the environment, not many units can be in- touching a picture images and illustrations of shops, restau- stalled in a building as a practical consideration. For rants, and so on. that reason, other sensors can be attached to the user’s device to enhance the positioning precision. (d) Users’ needs for cellular telephone services. (c) User interface. Although users utilized the smart phones prepared for us in the experiment, we found users’ preference use of the ser- In the questionnaire about an input to smart phone of the vice by which the indoor navigation service is also carried destination of the navigation in experiment, almost all par- out in a cellular telephone. The questionnaire results de- ticipants responded that they were able to understand the picted in Fig. 6(d) show that about 74% of participants an- swered that they would like to use such a system if the bea- ronments, some alternative positioning systems are pro- con receiver were miniaturized, or if the beacon receiver posed to use GPS in the locations where it is impossible to were unnecessary. receive a GPS signal from real satellites. The Pseudolite- GPS [3] and GPS Re-radiation [4] Systems are well known Conversely, it is a minority view that the user would not as alternative systems to utilize GPS. like to use the service in a cellular telephone. Therefore, although users’ needs for indoor navigation are high, for Pseudolite-GPS is a system using a transmitter that emits popularization of services, it is important that usage condi- the pseudo-GPS signal generated by simulations. However, tions also include the device configuration. the system has difficulty receiving signals at close distance and at distant locations from the transmitting antenna. Discussion Moreover, it is difficult to obtain correct time synchroniza- Based on the result of the experiment conducted at Yoko- tion accurately between the system and GPS satellites. The hama Landmark Plaza, we can discuss users’ needs for in- GPS Re-radiation System receives real GPS signals in open door navigation services. Results of the questionnaire to field areas, then forwards them with a cable, and transmits participants provided feedback that the users desire use of them to indoor environments. The system cannot be used in the indoor navigation service if the service can operate in areas that cannot also receive GPS signals outdoors. Addi- popular, practically used devices such as cellular telephones. tionally, it is necessary to install many GPS receiving an- tennas outdoors to raise the position tracking precision. Moreover, in the opinions of participants, users were unsure whether the walking direction matched the map aspect on Therefore, various non-GPS methods are often applied to the smart phone’s screen when the user was walking using indoor positioning systems. Active Bats [5] and Cricket [6] the indoor navigation service, after moving to another floor are techniques that can detect positions by receiving signals by escalator or elevator, after turning a corner, and so on. sent from transmitter devices. Typically, these systems in- crease the number of installed devices if the areas to recog- The disorientation occurred because the map screen display nize the user’s position become broad. aspect on the user’s terminal is fixed. Although the sense of direction depends on the person, in some situations, the In practice, RADAR [7], PlaceLab [8], EKAHAU [9], and users noticed no difference between a self-direction in the AirLocation II [10] are methods using the strength of radio real environment and the aspect of map displayed on the electric field of Wi-Fi signals. Improving these methods’ small screen. Also, the floor design of Yokohama Land- positioning precision is difficult because the electric field mark Plaza has similar architecture throughout the entire strength becomes unstable for changing of Wi-Fi radio area. To reduce loss of motion by disorientation, heading up transmission power. Ubisense [11] uses ultra wideband of the map through self-direction using an electronic com- technology (UWB) for positioning, but it is impossible to pass sensor in the cellular telephone must be effective, recognize the self-positions of users on their own portable which is our future work for this system. devices. We were acquired users’ opinions about the advantages and We have developed an indoor positioning system in consid- disadvantages of our indoor navigation system because eration of the shortcomings of the systems described above. evaluated the system by the several kinds of users. The sys- The system can operate on a portable information terminal tem was evaluated by two survey methods, that is, one is such as a cellular telephone by receiving radio beacon sig- description format and the other is group discussion. Al- nals from beacon devices installed in the environment. The though group discussion can hear the detailed opinions of system operates autonomously without server access; the the users, the survey method takes much time and energy. installed beacon devices can be driven using batteries at- In the feature, we would like to consider how to effective tached to the devices. Consequently, our system also pre- interview methods. serves user privacy. REATED WORK FUTURE WORKS In recent years, position information services such as navi- In our system, a guidance picture is displayed at a point gation services have received much attention in the context distant from a fixed distance. Therefore, we are developing of civil life, home life, industry, and so on. In outdoor situa- a technique for adjusting guidance points to resolve this tions, car navigation systems that specify self-position and problem. Moreover, to show guidance pictures with opti- which provide directions to the destination are useful as a mized timing on the screen, we consider an appending func- substitute for human navigators to such locations. Moreover, tion, which modifies it according to walking speed, which when advanced traffic information systems [1][2] are ap- is acquired in real time by another sensor. Furthermore, we plied to car navigation systems, it is also possible to show a are considering enhancement of the positioning precision to route with consideration of traffic and accident information attach other sensors in the user’s device [12]. Future sys- related to the present. tems might incorporate miniaturization of the beacon re- ceiver, integration as IC of the device, and integration with However, because the systems have difficulty receiving cellular telephones. signals of Global Positioning System (GPS) in indoor envi- 5 CONCLUSION 3. Stone, J., LeMaster, E.A., Powell, J.D. and Rock, S. M.: For this study, we have developed a system for an indoor "GPS Pseudolite Transceivers and their Applications", navigation service that is intended for implementation not Institute of Navigation National Technical Meeting, San only on smart phones but also on cellular telephones. Diego, California, USA (1999). Moreover, we presented results of an experiment in Yoko- 4. HNRRKIT: "Hanger Network GPS Re-Radiating Kit", hama Landmark Plaza for an indoor navigation system con- (http://gpsnetworking.com/) (2008-05-05). ducted using beacon devices. 5. Addlesee, M., Curwen, R., Hodges, S., Newman, J., Yokohama Landmark Plaza has a blow-by in the center of Steggles, P., Ward, A. and Hopper, A.: "Implementing a each floor above the ground floor, which means that it is Sen-tient Computing System", IEEE Computer Maga- difficult for users utilizing the service to recognize motion zine, Vol.34, No.8, pp.50-56 (2001). to another floor. In the building, many customers come to shops and restaurants; the people are passing through using 6. Priyantha, N. B., Chakraborty, A. and Balakrishnan, H.: the passages. We carried out the experiment to investigate "The Cricket Location-Support System", In Proc 6th this indoor navigation service in such a realistic environ- ACM Inter-national Conference on Mobile Computing ment. and Networking (ACM MOBICOM2000), Boston, MA, pp.32-43 (2000). In the experiment, we administered questionnaires to elicit 7. Bahl, P. and Padmanabhan, V. N.: "RADAR: an in- comments related to experiences of indoor navigation in the building RF-based user location and tracking system", commercial building. Consequently, we acquired useful IEEE Info-com2000, Vol.2, pp.775-784 (2000). opinions related to the navigation system and position in- formation services, such as the interface, its operation abil- 8. LaMarca, A., Chawathe, Y., Consolvo, S., Hightower, ity, and the screen information. Moreover, we confirmed J., Smith, I., Scott, J., Sohn, T., Howard, J., Hughes, J., the possibility of correct function for navigation service Potter, F., Tabert, J., Powledge, P., Borriello, G. and using indoor position information in this real environment. Schilit, B. "Place Lab: Device Positioning Using Radio Beacons in the Wild", In Proc. of the 3rd International ACKNOWLEDGMENTS Conference on Pervasive Computing 2005 We thank City of Yokohama, Mitsubishi Estate Company (PERVASIVE2005), Munich, Germany, pp.116-133 Limited, Musashi Institute of Technology, Incorporated (2005). Nonprofit Organization VMCY and Vitaminmama Com- 9. Ekahau: "Setting the Standard for Wi-Fi based Enter- pany Limited for their cooperation to our indoor navigation prise Visibility Solutions", (http://www.ekahau.com/) experiment. (2008-06-15). REFERENCES 10. AirLocation II: 1. Yamaguchi, M., Kitamura, T., Yoshida, T. , Horiuchi, T. (http://www.hitachi.co.jp/New/cnews/month/2006/07/07 and Takeuchi, K.: "Trial Application of the UTMS Dy- 13.html) (2008-06-23). namic Route Guidance Systems", In Proc. the 3rd Inter- 11. UbiSense: "Local position system and sentient comput- national Confer-ence on Intelligent Tutoring Systems, ing", (http://www.ubisense.net/) (2008-06-24). Orlando, USA (1996). 12. Sashima, A., Inoue, Y., and Kurumatani, K.: "Spatio- 2. Fujimoto, N., Namizuka, T., Takeuchi, K., Kitamura, T. Temporal Sensor Data Management for Context-Aware and Ogura, M.: "The Dynamic Route Guidance System Services", In Proc. of the International Workshop on of Nagano UTMS", In Proc. the 5th International Con- Advanced Data Processing in Ubiquitous Computing ference on Intelligent Tutoring Systems, Seoul, Korea (ADPUC2006) (2006) (1998).