Self-Driving Technology Enabling Smooth Avoidance of Pedestrians

Request Number REQ6370788
Due Date March 27, 2018
Request for Proposal Details
RFP Title
Self-Driving Technology Enabling Smooth Avoidance of Pedestrians
RFP Description

NineSigma, representing The Kansai Electric Power Co., Inc. ( ), seeks a self-driving technology that enables smooth avoidance of multiple moving pedestrians aiming to apply to a mobility for slow speed and short distance. The self-driving technology will be put to practical use for “iino”, a self-driven urban mobility service that Kansai Electric has been planning to provide. The system will be implemented on a low-speed vehicle and operated in demonstration projects in several cities in Japan


Kansai Electric Power is planning to provide “iino”, a self-driving urban mobility service. “iino” is a short-distance public transportation service with self-driven vehicles that circulate an urban area at low speed and allows passengers to get on and off freely without restricted by predetermined timetable or stops. Thus, “iino” will be provided as a unique transportation service, different from bus, taxi, and train. To accomplish this project, Kansai Electric Power seeks a potential partner with an excellent self-driving technology who can participate in the demonstration project, which has been planned jointly with multiple local governments.


In this project, it is assumed that vehicles will smoothly avoid surrounding pedestrians, bicycles and other obstructions on public roads and on sidewalks in a compact city. Therefore, the driving environment is more complex than expressways and control simply based on proximity sensors is not sufficient to achieve the smooth avoidance, regardless of the driving speed will be as low as 4 to 6 km/h. The client thinks technology other than proximity sensors, e.g. technology to recognize and identify objects, predict their behavior, and autonomously choose appropriate route based on analysis of camera images and/or other data on surrounding situation might be required.


Development of self-driving technology has been widely conducted, aiming at various purposes, from cars to last mile delivery, package delivery, and in-warehouse transportation. In order to identify an excellent potential partner interested in conducting a demonstration with local governments, Kansai Electric Power has issued this open request to a wide range of companies and universities around the world engaged in R&D of related technology.


Key Success Criteria

Final target of self-driving technology

A self-driving technology that enables prediction and smooth avoidance of the following objects on public roads and sidewalks in an urban area, without stopping the vehicle as much as possible

  • Moving pedestrians, bicycles, and cars
  • Obstructions, such as stopped pedestrians and objects

Note that the technology does not need to have been commercialized at this point.


Assumed environment for self-driving

  • Speed: 4 to 6 km/h
  • Driving area: Public roads and sidewalks in an urban area (an area where roads and sidewalks are not clearly separated)
  • Anticipated vehicle
    • Dimension: 0.9 m in width × 3.4 m in length
    • Shape: See photograph
    • Number of passengers: About 3 to 4
  • Available sensor: No limitation
  • Internet connection: Possible

A vehicle used for self-driving can be provided by Kansai Electric Power.


Anticipated Respondents

Anticipated respondents might be, but are not limited to, organizations who have the following technologies:

  • Self-driving technology for robots and vehicles used inside a warehouse, factory, on a street, etc.
  • Self-driving technology for last mile delivery
  • Self-driving technology for cars used on a public road full of pedestrians
  • Self-driving technology for personal mobility


Possible Approaches

Possible approaches include, but are not limited to, a combination of two or more of the following technologies:

  • Analysis of the environment with visible images and/or other sensors based on infrared rays, or sound waves etc.
  • Recognition and identification of objects, and prediction of their behavior based on analysis of images and/or other data obtained from the surrounding environment
  • Function improvement through a verification test, using machine learning and deep learning
  • Calculation of a route to smoothly avoid objects, based on motion prediction
  • Flexible judgement and routing function that does not rely on pre-programmed information (mapping technology)
  • Vehicle/ robot control technology compatible with the above-mentioned recognition/judgement function

Approaches not of interest

The following technologies are not of interest:

  • A partial solution
  • Simulation technology that is not aimed at actual implementation on robots or vehicles


Preferred Collaboration types
  • Joint Development
  • Consulting
  • Supply Agreement
Items to be submitted

Responses will use the Proposal Template which is available here and include the following items:

  • Overview of the technology
  • Characteristics and uniqueness of the technology
  • Development stage
  • Function to avoid obstructions smoothly
    • Sensor to be used
    • Principle of behavioral prediction
    • Principle of calculation of an avoidance route
  • Past achievements of self-driving test(s) for a car or robot
    • Size of the car or robot
    • Driving environment (e.g. presence of pedestrians and other obstructions)
    • Overview of the test(s)
  • Challenges for development and measures to overcome them
  • Cost needed to put the entire system in place and cost-reduction prospect
  • Request regarding collaboration (e.g. style, period, agreements, maximum time period of stay in Japan)
  • Past achievements (e.g. additional information supporting R&D capabilities, such as research papers and patents)
  • Corporate overview

Anticipated Project Phases or Project Plan

Respondents should submit proposals using the attached Response Template.


The client will review submitted proposals and possibly ask clarifying questions before selecting the most suitable candidates for collaboration. The client will select the best candidate(s) through evaluations. During the selection process, the client may execute non-disclosure agreements (NDA) with selected respondent(s), seek further information disclosure, and discuss specific development targets or potential opportunities.

The client will execute necessary agreement(s) with the selected respondent(s) and move to the advanced development phase. Specifics of any collaboration will be determined through consultation with the concerned parties.

Area of Interest
Request Priority