Customer Solution Case Study
/ SoftBank Mobile Supports Innovative Geo-Media Services Through Integrated Solution
Overview
Country or Region: Japan
Industry: Telecommunications Industry
Customer Profile
After making Vodafone [Japan] into a subsidiary in April of 2006, the SoftBank Group changed its name to SoftBank Mobile Corp. They became a hot topic as they rolled out a variety of services such as low-cost basic service plans and free calls between subscribers. They have had a major impact on the recent popularity of smart phones, and they are greatly expanding their subscriber base.
Benefits
· They are able to extract data from 3 billion records in 2–3 seconds
· Can create responsive geo-media services
· Able to construct an easy-to-use, visually attractive interface in a short time
· They have optimized scalability and costs by separating out how White Cloud and Windows Azure are employed / “At SoftBank Mobile...we announced the substantial goal of doubling the number of base stations we have in one year...I believe the system we are constructing now will be a valuable asset to our successors two or three decades down the road.”
Junichi Miyakawa, Director and CTO, SoftBank Mobile Corp.
SoftBank Mobile Corp. is pushing forward with a plan to double the roughly 60,000 base stations they have in one year as an opportunity to interface with their customers. To efficiently provide support for that effort, they are bolstering their existing geographic information system (GIS), constructing a variety of geo-media services (providing information related to location), including dynamic population data applications, mobile connections and information updates available to the general public. SQL Server 2008 R2 Enterprise Edition is the tool of choice for supporting their geo-media services database containing three billion records. For the backbone, which includes Web services for the public, they are using White Cloud, a cloud service from SoftBank Telecom, and Windows Azure, a cloud service from Microsoft.
Situation
Since entering the mobile phone market, SoftBank Mobile Corp. (SoftBank Mobile) has been the focus of much talk, with their energetic approach in providing innovative services, a product lineup with advanced features to meet customer needs, and serious development of smart phones.
On March 28, 2010, SoftBank Mobile issued a significant declaration to their users, entitled the “SoftBank Declaration of Wireless Improvement.” In it, they proclaimed they would eliminate dissatisfaction with wireless communication that customers complained about by “doubling” their base stations in fiscal 2010, which they had already been in the process of increasing since they introduced their third-generation (3G) mobile communication system in 2002. Junichi Miyakawa, Director and CTO of SoftBank Mobile, explained. “By March of 2010, we had about 60,000 3G base stations installed. Even in global terms, that is a large number. But wireless radio waves are living things, and they bounce off buildings, creating dead spots that are one of many factors frustrating customers. We reflected on the criticism we had received, and we created the ‘Declaration of Wireless Improvement’ to work on three measures aimed at improving the situation: doubling the number of base stations, providing free miniature base stations, and distributing free Wi-Fi routers.”
The “Declaration of Wireless Improvement” calls for a regional enhancement of radio waves by doubling the number of base stations, and providing free miniature base stations (known as femtocells) as part of a plan to improve service in locations pinpointed as having base-station radio wave interference or where SoftBank Mobile’s wireless 2 GHz waves do not easily reach. They are planning distribution of free Wi-Fi routers to ordinary businesses and users in three stages to promote the use of Wi-Fi for distributing the bandwidth load. A unique feature of this project is the open and flexible relationship that they maintain with the customer.
“The ‘Declaration of Wireless Improvement’ actually got its start with Twitter. Twitter is just one of the many means through which we build relationships with our customers. Our Representative Director [Masayoshi] Son followed up with a number of customers and received direct feedback, and it turned out that there were indeed many customers who were dissatisfied with their wireless situation. As I mentioned earlier, however, because of the nature of radio waves, this is not simply a matter of increasing the number of base stations. It takes a great deal of time to work out the details on locations where there is no signal, which may entail diagnosing such problems as the radio-wave interference ratio and poor reception underground.
So our idea was to invite users to send in connection problem reports from around the country, and we would then analyze each situation and pinpoint the exact location of the problem—that is, we decided to improve our wireless service based on customer feedback. To that end, we created an application for smart phones called the Wireless Signal Checker, which allows users to send us requests for improvement by simply pressing a button whenever they can’t get service.”
SoftBank Mobile continues to improve their service this way to this day. To support their improvement program, they have constructed GIS databases, and collected and analyzed base-station and area-simulation data.
In 2005, they integrated all of the GISs used by each department, creating their “integrated GIS-DB” which gave them the ability to comprehensively identify and analyze any mobile and broadband infrastructure data in the system, along with accompanying sales data. To encourage the use of that data, they created a Web-based “geo-media service” that can be accessed by any in-house employee, at any time. In addition, the use of the call center and technical support has assisted the selection of candidate locations for new base stations and other activities, which are still ongoing.
Nevertheless, the “Declaration of Wireless Improvement” laid out a mission to double the number of base stations in one year, a rigorous program given that it had taken them 10 years to get that many in the first place. To achieve their goal, SoftBank Mobile has refocused their reliance on data and developed mobile geo-media services to speed up the pace of installation. Through the smart phone application, they have established a system that compares information from the customer to data in the database, providing instant analysis and results, so that the company can follow up with a simple local test to determine whether the service needs to be improved. This has expedited the process of installing base stations.
President Son was very intent on SoftBank Mobile making every detail of the progress of this project open to the public, as the project got its start with a commitment to customers on Twitter.
While installing new base stations, they therefore began development of a publicly accessible geo-media service “Denpa Jokyo Pinpoint Kensaku” (Wireless Status Pinpointing Search) that will work in real time, showing their daily progress in wireless improvements to customers. Going public in October of 2010, they had constructed the system in a short time, designed with the availability and scalability needed for the heavy access from customers throughout the country that they expected—including the responses from Twitter. By opening a direct line of communication with their customers, responding and making the results public, and then asking for feedback again, they are engaged in “real-time marketing.”
The database that has supported these GISs and geo-media services is Microsoft SQL Server 2008 R2 Enterprise Edition, with Windows Azure Platform handling some of the resources for the enormous quantity of data.
Solution
Relational databases for easy geo-media services, easily handling 3 billion records
Since SoftBank Mobile began use of their stand-alone GIS in 2004, the database storing their accumulated and active data has been SQL Server. Kazuhisa Shibayama, General Manager of Area Solutions Planning Department Manager Area Planning Department, Mobile Solutions Division, looking back on the situation, said, “When we first started building the system, some people were concerned about whether SQL Server was the way to go. Since we had a lot of engineers on our staff who were experts at using Windows Server and SQL Server, we decided to construct a relational database (RDB) using SQL Server. Our data load continues to grow to this day, and SQL Server provides stable performance through each upgrade.”
Having confirmed the operational performance over the course of several years, we are adopting SQL Server 2008 R2 Enterprise Edition for use in the reconstruction of our system. We are running with 3 billion records of data.
The newly boosted geo-media services use Bing Maps for Enterprise to provide map data with an API (application programming interface). The interface is built with Microsoft Silverlight.
Shibayama explains the history leading up to the decision. “Initially, what we created for our geo-media services was broadband infrastructure information. When GIS middleware is used, you are using a technology known as ‘spatial joinit’ for plotting information on the map, but the processing is extremely time-intensive. At the time, we had five million users on ADSL, a volume which we could not handle as the data crunching for looking at Tokyo, for example, would take over half a day.
By employing ordinary Web technology, however, we successfully launched our geo-media services. The key factor was the standard support in SQL Server 2008 and later versions for spatial data forms (geometry and geography). Using that feature, we designed a relational structure with RDB tables to match the coordinates of buildings — which undergo relatively few changes — resulting in a boost of more than 10 times for Web page display times. Also, because the data is in an RDB, we can export to Excel, helping regular analysis go smoothly.”
Yuji Soga, Manager (Area Solutions Planning Department), recalled how they looked into Oracle Database during the construction, too.
“We did not really hesitate on the decision that much. With SQL Server, you get reasonable performance just with the default configuration, but with Oracle, there is a huge amount of tuning required for the parameters. Not only that, but we received regular support from Microsoft headquarters and reviews on our system architecture employing SQL Server and Bing Maps. Given the continuous stable performance we had with SQL Server, plus the fact that installation and maintenance costs are about one-third those needed for Oracle, it was an easy decision to go with SQL Server 2008 R2.
To give you an example, ‘Denpa Jokyo Pinpoint Kensaku’ (Wireless Status Pinpointing Search) is predicted to generate quite a few transactions. Not only our 20 million existing customers, but potential customers who are looking at switching from other carriers use this feature to check the service status at their locations 24 hours a day, 365 days a year, whenever they feel like it. Since they’re waiting for a response from our database, the demand on the system is really high, but we expect that SQL Server 2008 R2 will handle it without any problem. On top of that, Bing Maps is based on an API so it doesn’t have to be held within the database, and the costs are about 70 percent less than purchasing map information on a project basis as we had in the past. Another thing is that a Bing Maps controller is included in Silverlight, making interface development easy.”
The new system built with this “integrated GIS-DB” is accessible in-house as a “geo-media service for anyone to use anytime, anywhere” to meet the age of the smart phone.
A typical feature is dynamic population estimation data, developed by SoftBank Group based on data that has been continuously accumulated. Not limited merely to residents, new students and people joining the work force, this is the first collection of data in the world to include shoppers, recreationists, tourists and more by time series (by hour, day of the week, month or season). Using this data, you can make detailed assessments of a dynamic population on a dynamic screen. In addition to identifying semi-fixed features such as how buildings are laid out, it is possible to determine and analyze variation in the number of accesses to a base station down to the hour, making it possible to create a more detailed wireless improvement plan.
In order to curb capital investment as far as possible, they are utilizing the public cloud Windows Azure Platform for dynamic population estimate data and other information, which they are making available for public access. This method of applying appropriate technologies to each area of the system optimizes the construction costs for the system as a whole.
The heart of the wireless signal improvement program that SoftBank Mobile is currently promoting encourages customers to transmit information, an element of the program that has resulted in some 4.5 million submissions as of October 2010 through their smart phone application “Wireless Signal Checker” alone. In addition to providing SoftBank Mobile with continuous data on outdoor wireless status, Wireless Signal Checker has a save function that allows users to register locations outside of the service area, which will be transmitted at a later time.
For their newest geo-media service, SoftBank Mobile has established a wireless signal improvement policy for comparing the data users send in to the wide range of data the company has collected, including dynamic population estimate data. If the wireless signal interference status at a location needs to be known, a manager pinpoints the location on the geo-media service map on a computer, and sends a work request to a worker in the field. The worker confirms the request using the geo-media services screen on a Windows Phone and physically moves to the location. The worker operates a dedicated application, which collects more detailed information and sends it to the main office, which will be useful in the final improvement plan. The operation of the application essentially requires the simple click of a button, and does not require a special technician. This simplicity makes it possible to process more locations, more quickly.