Keywords

1 Introduction

No matter how well travelers have planned prior to departure, they must invariably modify their travel plans after arrival at their destination. For example, they may find themselves more tired than expected and hence decide to cancel planned visits to certain spots. Plan change is a reality that travelers must confront, especially towards the end of their tour.

Cyber-physical systems (CPS) seek to provide users with optimal control of the world they correspond with by modeling physical space in cyber space, coupled with the use of related databases. More than big data systems, social CPS is the operating system of urban society. It provides a user environment that supports the agency of people in decision-making. Social CPS is increasingly necessary for building sustainable, safe, and secure urban societies. The prerequisite basic technologies are maturing rapidly. Remaining efforts include opening data silos maintained by both the private sector and government, and analyzing massive, complex data that cannot be fully described with a single monolithic model. Tantalizing research and development challenges exist in relation to Social CPS.

This paper proposes a new service model for instant travel assistance that uses input from locals. We consider the model an example of Social CPS that aims to develop a framework to collect traveler needs based on sensor data that reflects microscopic personal-scale events, in a similar manner to crowdsourcing, and organizes locals to collect knowledge and recommendations that can support travelers.

2 Background

Travelers may have to modify their initial plans en-route, and various factors can require this. Weather is one of the biggest influences. Travelers can also find that their interests change during a long journey. Even if circumspect travelers can prepare alternative options to deal with such uncertainty, it is difficult to develop alternatives for all eventualities prior to their travel. Plan change thus is one certainty that travelers confront, especially towards the end of their itineraries.

We assume that a problem exists in that it is difficult for travelers to get useful information or effective support to help them modify their plans. This problem may cause not only the loss of opportunities for travelers to experience local interesting places and things, but also the loss of opportunities for local businesses to sell products and services.

2.1 Travel Assistance

Many travel assistance services have been proposed, using several approaches. One approach is online travel guides, such as Lonely Planet and AAA TourBook guides. Another approach is to supplement travel guides with user comments and reviews, in the manner of TripAdvisor. Social networking sites like Twitter and Facebook are a natural source of destination information. Web searches like Google must be frequently used. We suppose that, for various reasons, such existing services are insufficient to support travelers, especially on-site. One reason is limited time. In contrast to the situation when preparing for travel, once on-site travelers have limited time to research and plan. Another reason is limited tools. Once at their destinations travelers can only use their own smartphones for navigation. Hence they need to efficiently find things of interest using only their smartphones and while working under time restrictions. The problem thus is one of content accessibility. We also suppose that another essential problem involves insufficient local content. Generally, we often confront a lack of listed candidates when running searches for local restaurants. This is a problem of sparseness of content. Additionally, both social networking site content and “recommended” content on commercialized media may have low credibility. We must consider not only how to efficiently provide credible information to users, but also how to incrementally produce useful local content.

2.2 Mobile Applications

Numerous network services for mobile users with location data have been created, and some, such as foursquareFootnote 1, are becoming popular. Location information is usually given in the form of geographical coordinates, i.e. latitude and longitude, a location identifier such as a facility ID entered into geographical information services (GIS), or a postal address. Google has launched Google PlacesFootnote 2, which gathers place information from participating networkers and delivers it via Google’s website and application programmable interface (API). Google thus tries to obtain information on activities in the real world, information it lacks despite being the omniscient giant of the cyber world. Google already uses its own physical resources to capture real world information. For example, it gathers landscape images for the Google Street View serviceFootnote 3 using its own fleet of specially adapted cars. However, capturing and digitizing facts and activities in the real world is generally very expensive beyond a superficial level, such as capturing photo images associated with geographical information. Although Google Places may be a reasonable solution to gathering information in the real world, these is no guarantee it can become an effective and reliable source reflecting the real world.

Existing social information services, such as Facebook and Twitter, are expanding to attach location data to user content.

2.3 Social CPS

To capture situations and help solve problems in the real world, research on CPS has recently become increasingly important. CPS is a promising new class of systems that deeply embed cyber capabilities in the physical world, either on humans, infrastructure or platforms, to transform interactions with the physical world [5, 10]. CPS facilitates the use of information available from the physical environment. Advances in the cyber world, such as communications, networking, sensing, computing, storage, and control, as well as in the physical world, such as materials and hardware, are rapidly converging to realize a class of highly collaborative computational systems that rely on sensors and actuators to monitor and effect change. In this technology-rich scenario, real-world components interact with cyberspace via sensing, computing and communication elements.

Fig. 1.
figure 1

Social CPS

Full size image

Social CPS focuses human aspects in the parallel world because humans are not only subjects that exploit such systems but also objects that are observed and affected by the systems. Information flows from the physical to the cyber world, and vice-versa, adapting the converged world to human behavior and social dynamics. Indeed humans occupy the center of this converged world since information on their operating context is the key to the adaptation of CPS applications and services. The concept of social CPS is shown in Fig. 1(a).

2.4 Crowdsourcing for Civil Problems

The term “crowdsourcing” was coined by Jeff Howe in 2006 [7] to describe the act of taking a task traditionally performed by a designated agent and outsourcing it via an open call to a large but undefined group of people [8]. Crowdsourcing can involve peer-production, but is also often undertaken by solitary individuals [6].

The concept of smart cities can be viewed as a recognition of the growing importance of digital technologies to competitiveness and sustainability [11]. The smart city agenda, which sets ICTs strategic urban development goals such as improving the life quality of citizens and creating sustainable growth, has recently gained considerable momentum. In smart cities, collaborative digital environments facilitate the development of innovative applications, starting from local human capital, rather than believing digitalization can transform and improve cities.

Tools such as smartphones provide the opportunity to facilitate co-creation between citizens and the authorities. Such tools have potential to organize and stimulate communication between citizens and the authorities, and allow citizens to participate in the public domain [4, 12]. One example is FixMyStreetFootnote 4, which enables citizens to report broken streetlights and potholes [9].

Incidentally, to capture situations involving a town, such as local events or the feelings of locals and visitors, the author believes that it is not enough to collect tweets and behavior logs of locations, simply because the number of geotagged tweets is limited. In a previous work by the author [1, 3], only 1 % of LBS users posted microblogs while strolling in town. Therefore, for microscopic analysis of town situations in small resolution of time and space, more information sources that reflect pedestrian behaviors and emotions are needed [2]. Unconstrained crowdsourcing approaches will not succeed automatically and social standards like trust, openness, and consideration of mutual interests must be guaranteed to make citizen engagement in the public domain challenging.

3 Methodology

In the case where a traveler’s initial plans change on-site, we assume that it is difficult for travelers to get useful information or effective support to modify their plans. To solve this problem, we propose a new service model for the provision of instant travel assistance. The service targets on-site travelers and aims to help travelers make instant plans by providing them with plan proposals by locals.

Fig. 2.
figure 2

Service process flow

Full size image

3.1 Service Scenario

Travelers can send a plan request that incorporates various important conditions, such as an end point (destination and time), preferences (including those regarding cuisine, historical spots, etc.), and group information. They do this by completing the web form (Fig. 2. (1)), (Fig. 3(a)).

When the service receives the request (Fig. 2. (2)), it is delivered to cooperating locals (Fig. 2. (3)).

According to the request, locals make their own plans and register these with the service by the deadline (Fig. 2. (4)). If a plan includes a destination not registered in the service database, that destination will be added at this time.

On the arrival of the deadline, a plan proposal is composed using plans collected from locals and an email that includes a URL pointer to the plan proposal is sent to the requester (Fig. 2. (5)).

The requester can get the plan proposal, including plans proposed by Locals, by clicking the URL pointer in the email (Figs. 2. (6), 3(b)).

Fig. 3.
figure 3

Snapshot images of the service

Full size image
Fig. 4.
figure 4

Snapshot images of the service

Full size image
Fig. 5.
figure 5

Examples of Spot Information

Full size image

3.2 Features

This service model contains the following features:

  • Sending the request is simple and quick.

  • Plans are not only ready-made but also made-to-order by locals who know the area well.

  • Previously created plans and content are registered in the service DB and can be reused; the volume of local content thus gradually grows.

Figure 4 shows examples of plans created by locals. Each plan (Fig. 4(a)) includes the planner’s profile with a photo image, some general comments, and finally the details of the plan. In contrast to plans by locals, web search plans are simple; that is, they merely comprise the result of a web search related to the request.

3.3 User Functions

Typical Plans. Users can browse typical plans prepared by locals.

Spot List. Users can look up a list of already registered spots. They can also sort those spots according to distance from their current location.

When they select one spot on the list, spot information is shown. Figure 5 illustrates examples of spot information. Spot information consists of short descriptions of the spot, a map and business hours, as well as a “Like” button that users can use to express interest.

3.4 Sensing Functions

User Data. The service collects the following user attributes:

  • gender

  • generation

  • zip code

The service collects users’ demographic attributes on their first access.

Location. The service is accessed via web browser. When users post their plan request and search nearby spots, the browser sends the service their current location.

4 Survey

4.1 Overview

We conducted a preliminary survey of the proposed model at the beginning of March, 2014 in Matsuyama, Japan. We asked 26 actual travelers who carried their own Android smartphones to participate in the survey. The travelers were asked to use our proposed service at least once; that is, they were asked to request a plan proposal. Doing this required participants to install our smartphone application, “Ryoimp”Footnote 5, to allow collection their behavior logs in background mode.

Participants also had to compete an online survey after completing their travel.

4.2 Results

Sixteen participants completed the software installation and registration. Unfortunately, only seven participants sent plan proposal requests while eight answered the online survey.

Online Survey. We requested monitors to answer questionnaires on the following:

  1. 1.

    about plans provided by locals

    1. (a)

      expertise, consideration of preferences, and usefulness

    2. (b)

      comparative acceptance of plans created by locals and plans obtained from web search

    3. (c)

      whether the spots in provided plans were visited

  2. 2.

    acceptance of agent reservation service

  3. 3.

    services used during travel

    1. (a)

      useful internet services

    2. (b)

      useful information sources.

Expertise, Consideration of Preferences, and Usefulness. Two in five agreed that the plans provided by the service displayed expertise and consideration of requesters’ expressed preferences. Three acknowledged that the service was useful. None of the monitors who responded to the online survey gave negative ratings.

Comparison of Acceptance Between Plans Provided by Locals and Web Search Results. When examining the credibility questionnaire results, we found plans provided by locals to be more credible than those resulting from a web search (Table 1).

Table 1. Credibility: “Is the plan by {locals \(\mid \) web search} credible for you?”
Full size table

Four out of five participants who completed the survey and used the service at least once answered “real locals with viewable profiles providing customized plans is credible”. We suppose that these results support our assumption, namely that plans provided by locals with viewable profiles can be accepted as credible and preferable.

Actual Traveler Behavior. Table 2 shows the behavior of travelers who used the application, received customized plans and provided behavior logs. Unfortunately only a few participants provided such feedback, but four out of six recommended spots were accepted by these participants. We regard responses of “satisfied” and “want to visit” as indicating acceptance.

Table 2. Recommended spots and traveler behavior
Full size table

Figure 6 illustrates an example of traveler trajectory.

Fig. 6.
figure 6

Examples of a traveler’s log

Full size image

5 Conclusions

This paper proposes a new service model whereby travelers receive instant travel assistance from locals. Although we have not collected enough data, the questionnaire results support our approach. For example, participants found information from locals extremely useful, and regarded plans provided by locals as more credible than content obtained through a simple web search.

The authors are continuing to develop this service model, and experiments are being planned to evaluate its effectiveness.