Skip to content

Home

Effective policy making support 'Policy Twin'
This is the policy making support technology for local governments and private companies that can simulate the effects of services for residents in advance, taking into account individual residents, without relying on past experience or intuition.

Challenge in Providing Services to Residents

(画像を表す適切なテキスト)
In the area of services for residents to solve social issues, various policies have been maked, examined and provided by both local governments and the private companies. However, it has been difficult to calculate the effects before providing the service, and we have had to rely on past experience and intuition.

Policy Twin

(画像を表す適切なテキスト)
AI analyzes and digitizes various policy documents to optimize conditional branches. By utilizing the knowledge of empirical econometrics and conducting prior verification of residents' behavior choices through digital rehearsal, it is possible to efficiently search for optimal services for residents. This is achieved by calculating effects, comparing references to other cases, and simulating changes of policies.

The benefits of Policy Twin

  1. Proposal of policies based on the concept of resource allocation in empirical econometrics
    • Easier to make policies that take into account the population structure and services of each local government
  2. Simulating the effects in advance to streamline policy making
    • Reduces time spent on policy making and facilitates social consensus building

Demo Click here to try the demo
Please read the manual for the operation.

Technical overview

Target Industry/Users

  • Policy making and implementation departments of government and local governments
  • Strategy, planning and consulting departments of private companies

Challenges in Target Industry and Operations

  • It is necessary to improve policies that determine which services are provided to whom within the limited social resources, and to maximize the effects.
  • Since it is difficult even for experts to calculate the effects of policies in advance, they often rely on the intuition and experience of those in charge.
  • As stakeholders such as local governments, private companies, and residents differ in their opinions, it takes time to reach consensus.

Technical Challenges

  • Digitize policy documents (convert to machine-readable form)
  • Generate candidates for more effective policies based on digitized current policies
  • Calculate how the target indicators will change when the policy candidates are implemented

Solutions

  • Convert policies to machine-readable flowchart format using large-scale language models, etc. (Patent Pending)
  • Reconfigure new policy flowchart candidates by combining parts of the flowcharts of successful policies. (Patent Pending)
  • Calculate target indicators by simulation using the policy flowchart candidates. (Patent Pending)

The benefits of Policy Twin（Detailed version）

  • Propose new policy candidates by reconfiguring widely collected policies automatically.
  • Verify in advance by simulating on policy candidates and calculating target indicators such as effects and costs.
  • Find the policy that maximizes the effect by comparing the difference between multiple policy candidates.

Fujitsu's Technological Advantage

  • Support efficient policy making through the reconfiguration and simulation of policy candidates based on machine-readable policy and performance data.
  • Facilitate consensus building among various stakeholders and implementation in society by showing the basis of the proposed policies.

Use Cases

  • Policy maker
    • Analyzing existing policies and developing new ones to address various social and business issues.

Case Studies

  • Multiple duplicated medications in health services
    • Background
      • It is a problem that people visit multiple medical institutions and receive the same medicine or take many kinds of medicine.
      • Not only do they not get the benefits they deserve, they also risk side effects and increase medical costs due to extra prescriptions.
      • Local governments use medical history data on residents to identify those who are eligible for the program, and public health nurses and others visit residents to provide guidance and encourage them to change their behavior.
    • Challenges
      • Each local government carries out the challenge for health promotion of visiting guidance, etc. for the elimination of the double ingestion.
      • However, few local governments are able to produce sufficient effects, and improvement is required.
      • It is difficult to formulate effective policies because determining who should receive what kind of guidance is complex.
    • Resolved with Policy Twin
      • Convert policies of multiple local governments into a machine-readable flowchart format.
      • Create new policy candidates based on the policies of other successful local governments and propose them as new policies that can solve the problems of the target local governments.
      • It is possible to select the policy that maximizes the effect by presenting the target indicators when executing multiple policy candidates and comparing the difference in the effect.

Technical Trial

Documents

Document Title Explanation
Web operation manual Demo page operation manual
Data generation manual Description of the data used in the demo

Demo Click here to try the demo
Please read the manual for the operation.