|Company Name||DATAFLUCT, Inc.|
DATAFLUCT is a data science startup studio that has continuously created new businesses since its founding in 2019 by using previously unused data to create value for companies and society, based on our vision of “turning data into business.” We use a wide range of analysis results, such as satellite image data to location information and POS data. We are skilled at building algorithms that can transcend industries. We provide DX support for various companies by utilizing the knowledge that we have accumulated in the development of various in-house services, from food distribution to the real estate sector. We also aim to contribute to SDGs by using data. Moreover, we are actively working on developing new businesses that can strike a balance between fiscal and social contributions. (2019 JAXA Venture Certified Company)
This project aims to develop DATAFLUCT area management, an active visualization platform that integrates and visualizes all data related to area management to support decision-making and the PDCA cycle. Activity in an area is defined in terms of “human traffic,” “emotions,” and “economics.” Moreover, it is shown on a map as GPS/location data (volume/migration), payment data, and social media data, respectively. It can be adjusted to the area management guidelines for each company and supports the import of in-house data. The program features data export and in-house URL sharing functions and accelerates communication among responsible parties for regional revitalization.
For about 1 km around Shibuya station, we visualized the following to analyze the causal relationship among variables and develop our minimum viable product (MVP).
There are 18 defined area segments within about 1 km from Shibuya station. The major commercial complexes in the area include Shibuya Stream and Shibuya Hikarie, among others.
We procured data for 2 months (i.e., December 2019 and December 2020) and used it to confirm the transformation of medium to long-term activity in the area before and after the COVID-19 pandemic and short-term activity caused by events.
（Other relevant data）
We created a visualization of the volume of human traffic according to 5 attributes in each area segment/building: “worker,” “resident,” “worker/resident,” “visitor,” and “passerby.” The volume of human traffic in the area can be visualized according to their place of residence, whether that is “the 5 districts of Shibuya Ward,” “ward/city of Tokyo,” “the 3 prefectures in the metropolitan area,” or “other prefectures.”
The migration amount for each area segment can be shown concretely according to the time period. It is used to visualize the volume of human traffic from the primary movement destination to secondary and tertiary destinations during the specified period. Time-series changes in the volume of human traffic heat map for the entire area can be visualized by date and time to show the amount of migration in 25 m grid units. We have also attempted to visualize the subsequent movement grid in chronological order for human traffic at a specific commercial facility at a designated time.
Cash register data in commercial facilities can be visualized in the following tenant categories: “Japanese restaurant,” “Italian/French/Western restaurant,” “Chinese/Ethnic restaurant,” “Cafe,” “Other.” The program can also be used to confirm the causal relationship with changes in the volume of human traffic.
The program provides insight for causal analysis by visualizing changes in human traffic during a given period according to “attributes,” “place of residence,” or “by hour.” To analyze changes in the volume of human traffic during a given period, we can show keywords on Google Trend that are quickly increasing in popularity using DATAFLUCT’s original algorithm (e.g., increase in the flow rate of people on December 26, 2020, → trend word “Haikyu Exhibition”)
In this proof of concept, we successfully implemented various analysis items using feedback from task force companies for their usefulness in the decision-making processes of area management. Using these analysis items, we were able to obtain key findings in areas such as the decrease and recovery of the volume of human traffic before and after the COVID-19 pandemic began (up to a 30% increase/decrease when comparing December 2019 and December 2020), the migration status of visitors, and the volume of human traffic in events, and the relationship between factors and trend words. In the future, we will experiment in areas such as Umeda in Osaka and Yaesu in Tokyoto to develop universal solutions that can be adopted to each companie’s area management policy. Additionally, we plan to develop a solution that can quantitatively and objectively analyze area management data from various perspectives by expanding the types of data, POS and SNS. In June, we plan to provide this software as a commercial solution to real estate companies, governments, and event-related companies, as well as to our partner companies in this Proof of Concept. For inquiries, please contact Yamada at firstname.lastname@example.org
We will send you the latest events information and articles
or notify opening of application for each Onlab's Program, and more.