Tsukuba field trials

This blog is dedicated to BigClouT field trial to be conducted on the behalf of Tsukuba City and NTTE.

Flag_of_Tsukuba_IbarakiNTT East

Tsukuba City is in interested in using BigClouT platform for developing sightseeing and current transportation options.

The BigClouT project just started July 1st 2016, here is the timeline for defining the first set of use cases and field trial.

  • September – December 2016: Use cases definition
  • January 2017 – June 2017: Field trials preparation
  • July 2017 – December 2017 : Field trials deployment and evaluation

If you want to share ideas or be involved in the definition and validation process, please contact us!

~~~~~~~~~~~~~~~~~~~~Field trials~~~~~~~~~~~~~~~~~~~~

* Use Case 1: Provide tourism, traffic and environmental information in real time to visitors

Description

NOTE: This trial is still in a planning phase. We are considering integrating this trial plan into Trial 2 at Tsukuba City (Participatory Sensing App for Foreign Visitors) and to have a combined trial in the year of 2018.

Tsukuba is famous as a scientific city, where many research institutes are located, as well as a sightseeing place. For this reason, there are a lot of Japanese and foreign tourists to Tsukuba. Anecdotal evidence informs us that some visitors are not satisfied with their visit e.g., cannot climb the Tsukuba mountain due to bad weather, lack of information about popular scenic spots etc. One main reason is poor planning. Therefore, this trial aims to improve visitors’ satisfaction by recognizing their behaviour and:

  • Provide real time and optimal tourism information or tour route by recognising their behaviour and predicting environmental information (ie., real-time warning about routes with traffic congestions/shops with long queue/dangerous weather condition/… based on current location info).
  • Recommending to tourists useful information, such as the popular scenic spots, and restaurant in the city so that they will stay longer or revisit Tsukuba again (ie., recommending scenic spots to foreigners based on profile info).

We have had discussions with city officials in Tsukuba City regarding this trial plan, and have their feedback regarding possible information sources and stakeholders, which are already integrated in the trial plan. An overview of the trial technology framework is shown in the figure below. Data is collected from devices, passed through BigClouT cloud infrastructure and pushed to edge devices such as digital signage boards.

Scenario

  1. Collect attribute information such as purpose for visiting to Tsukuba, transportation, gender, generation, hobby and tastes.
  2. Collect behaviour history of the visitor. Collect visitors’ satisfaction rate for purpose of the visit and other demands. Collect environmental information (weather, temperature and so on).
  3. Analyze the trends of visitors’ behaviour according to their attribute and environmental information and the trends of satisfaction rate.
  4. Collect real time weather information, Social Network Service comments, traffic and facilities’ congestion situation, and visitors’ location information.
  5. Select the best tourism, transportation or restaurant information for the visitor according to their attribute and real-tie location information by data of 3 and 4. Predict this information for the future by accumulated information.
  6. Deliver this information to visitors.
  7. Get feedbacks about information delivered by questionnaires and so on.  Improve recommendation accuracy based on feedbacks.

Goals and Stakeholders identified

The technical goal is to assess the usefulness and effectiveness of the BigClouT architecture in the context of real-time assistance of tourists using big data and predictive analysis technologies. The user focused goal is to improve the quality of the visit for participating visitors. A subsidiary goal is to better understand needs of visitors.

Several stakeholders have been identified for this use case:

  • Tsukuba city
  • Tsukuba Tourism and Convention Association
  • Tsukuba City Bussankai
  • Tsukuba International Congress Center
  • Tsukuba Science Tour Office
  • Metropolitan Intercity Railway Company
  • Tsukuba Scenic Railway Co., Ltd.
  • Kanto Railway Co., Ltd.

Ongoing engagement

Details of ongoing engagement as initial work has begun.

This trial is still in a planning stage. We have had several meetings with city officials in Tsukuba City, and received feedback regarding the trial plan, such as possible information sources, use cases, etc. We have reflected the feedback to the trial plan.

* Use Case 2: Grasp status about foreign visitors to Tsukuba and provide concierge service to them

Description

Tsukuba city is one of the most famous scientific cities in Japan with more than 100 research institutes and universities. For this reason, many domestic/international academic meetings, including international conferences, are held in Tsukuba. Additionally, Mt. Tsukuba is a famous sightseeing spot which receives many visitors from Japan and overseas each year.

According to city officials in Tsukuba, they do not have comprehensive information regarding foreign visitors, such as the number, gender, country, popular visiting places, etc. and they do not always know the problems experienced by foreign visitors.

To cope with this problem, in this trial, we develop a participatory sensing system for foreign visitors using smartphones to collect problems experienced by the users during their stay in Tsukuba. The application allows users to submit a problem along with a photo and descriptive texts. In addition, for submitted problems, the system allows volunteers to offer real-time responses in terms of texts, which are considered to be useful for the visitors to solve the problem. The objectives are as follows: 1) to collect problems experienced by foreign visitors in Tsukuba, 2) to make real-time assistance by volunteers to the visitors through the system, and 3) to collect basic information about foreign visitors in Tsukuba, such as profile, age, gender, living country, visited places in Tsukuba, etc.

The following shows some screenshots of the developed smartphone app.

Scenario

  1. Collect foreign visitors’ attribute information such as purpose for visiting to Tsukuba, transportation, gender, generation, hobby, tastes and language.
  2. Collect needs and problems of foreign visitors to grasp their situation.
  3. Collect behaviour histories of foreign visitors.
  4. Collect and accumulate information that would solve the needs and problems.
  5. Prepare for the most suitable information for the demands (problems) of foreign visitors to Tsukuba from collected and accumulated data.

Example:

・Instruction to the local destination (for any vague needs such as “I want wine!”).

・Guide the instruction to the area including downtown, not a specific shop (e.g.: an area with many sushi restaurants).

・Place (or direction) that is easy to get a taxi.

・Emergency contacts (hospital, police).

・Restroom map (degree of cleanliness).

・Getting information to avoid dangerous areas.

・Waiting time of restaurants and List up of the restaurants by waiting time.

  1. Deliver answers for the inquiries from foreign visitors according to their attribute and real time location information. Deliver additional recommended information to the person who inquires
  2. Get feedbacks about information delivered by questionnaires and so on. Improve recommendation accuracy based on feedbacks.

Goals and Stakeholders identified

The goals of this trial are as follows: 1) to collect problems experienced by foreign visitors in Tsukuba, 2) to make real-time assistance by volunteers to the visitors through the system, and 3) to collect basic information about foreign visitors in Tsukuba, such as profile, age, gender, living country, visited places in Tsukuba, etc.

Several stakeholders have been identified for this use case:

  • Tsukuba city
  • Tsukuba Tourism and Convention Association
  • Tsukuba City Bussankai
  • Tsukuba International Congress Center
  • Tsukuba Science Tour Office
  • Metropolitan Intercity Railway Company
  • Tsukuba Scenic Railway Co., Ltd.
  • Kanto Railway Co., Ltd.

Ongoing engagement

  • We have had a small trial with students in University of Tsukuba to collect initial feedback. We made use of the feedback to refine the user interface of the smartphone app.
  • We have had several meetings with the city officials of Tsukuba City to discuss the trial plan (1 to 2 times each month for 7 months before the trial in addition to about 20 meetings with the stakeholders). According to the result, we put posters in various places in the city, such as station, conference halls, University of Tsukuba, etc. In addition, we put flyers in different places in the city. On the posters and flyers, we put QR code whereby users can install the smartphone app in their own devices. In total 50 posters and 8,000 flyers were distributed at the places below;

1) Tsukuba City Hall 2) TSU campus 3) Stations in Tsukuba (4 stations) 4) Mt. Tsukuba (4 places) 5) bus (3 routes) 6) Tsukuba International Congress Center 7) Souvenir shops 8) hotels (19 hotels) 9) Housing for Foreign Researchers (2 places) 10) Tsukuba international association.

  • We had realized that stakeholder’ support was necessary because it would be very hard to recruit participants from foreign visitors. Tsukuba University and Tsukuba city officers have therefore had more than 20 meetings with the stakeholders to ask for cooperation and to get feedback to the trial plan. The trial plan was refined based on this ongoing feedback process.

Lessons learnt

Details of lessons learnt from the ongoing engagement, from feedback from stakeholders and from any initial trial activities

– We are evaluating our initial data from the small scale trial (Nov-Dec). One clear lesson is related to recruitment which was lower than hoped. One major reason is the language barrier; i.e., we assumed English as the basic language in the app and advertisement, whereas it turned out that most of the foreign visitors in Tsukuba City are Chinese. To recruit more participants, we need to support different languages so that people from different countries can participate readily.

– Thanks to the supports from the stakeholders, we have very valuable feedback from them after the trial which is outlined below. We are analyzing this feedback and developing lessons for the next trial.

  • Kanto Railway Co., Ltd.

* Many foreigners (mainly Asians) were visiting Mt. Tsukuba and they took the flyer.

* It is useful to know that some people posted the problems about a bus. We would like to cooperate on next field trial.

* Problems about a bus were not many. One of the reasons may be our staffs stood to guide visitors at the bus terminal in November since many people visit Tsukuba in November.

  • Tsukuba Tourism and Convention Association

* Condition for present exchange, which is to post 5 problems and answer the questionnaire, was too severe.

* The application should be available not only in English but also in Chinese or other Asian languages since 50% of foreign visitors to Tsukuba is from China.

* The application should be available for iOS because there are many iPhone users.

  • Tsukuba international association

* Many foreigners who live in Tsukuba city use iPhone. So support for iPhone is necessary

  • Tsukuba Scenic Railway Co., Ltd.

* Have distributed all flyers.

* Most foreigners visiting Mt. Tsukuba was Asian people.

  • Tsukuba International Congress Center & Tsukuba Science Tour Office

* Condition for present exchange was too severe. It would be better if users can get present at hotel they are staying.

* The number of the problems posted seems reasonable since only a few international congress / science tour for foreigners were held in November. Problems posted seem to be useful.

* It would be better if the target of HukuRepo is expanded to foreign residents in Tsukuba (not only visitors).

* The application should be available in Asian languages.

* Should co-operate with research institution in Tsukuba because they hold international meetings.

* The application should be available for iOS because there are many iPhone users.

  • Tsukuba City Bussankai

* It seems many foreigners who live in Tsukuba City or who visit the international congress centre understand Japanese. And visitors from overseas are mainly Asian people who don’t speak English. These may be the reason why the number of the users of HukuRepo, which was only available in English, was not many.