Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Using Automated Vehicle Data as a Fitness Tracker for Sustainability

Xia Wang, Sobenna Onwumelu, Jonathan Sprinkle

TL;DR

The paper tackles how to help drivers contextualize driving sustainability by leveraging on-board CAN-bus data from ADAS/AV-equipped vehicles to produce a post-drive dashboard. It introduces a sustainability dashboard that computes and visualizes indicators for safety, fuel efficiency, and comfort, using metrics such as headway, TTC, a CAN-based fuel-consumption model, and acceleration-jerk based comfort. Key contributions include a defined data format for driver-indicator interaction, a low-latency CAN-signal pipeline, and a user-friendly interface with trend and comparison capabilities to foster trust and acceptance of autonomous driving assistance. The work demonstrates how exposing under-the-hood vehicle data in an interpretable format can support driver engagement, guide iterative controller development, and enable data-driven improvements in AV safety, economy, and ride quality.

Abstract

This work describes the use of on-board vehicle data from cars with advanced driver assistance features as a trip summary, with the goal of helping drivers contextualize their driving habits in terms of sustainability. The approach is similar to recent advancements in fitness tracking apps, which leverage smartwatches and other wearable devices to characterize activities during a workout or as part of daily fitness monitoring. Instead of adding new vehicle sensors, the data used for this work is from on-board driving data, namely, signals decoded from the vehicle's Controller Area Network (CAN) bus. With the deepening research of automatic driving technologies, Autonomous Vehicles (AVs) have gradually entered the consumer field, and more users are benefiting from the convenience and safety assistance provided by driving assistance and autonomous driving. However, various technical obstacles persist due to the complex environment, the non-communication of technologies, and users' trust. We propose indicators for evaluating the key characteristics of each drive, to facilitate drivers' familiarity with advanced driver assistance systems and to allow them to consider how different driving styles affect sustainability metrics. Further extensions will allow users to add feedback as part of the driving summary, laying a data foundation for future controller iterations based on real driving data and the attitude of drivers towards it.

Using Automated Vehicle Data as a Fitness Tracker for Sustainability

TL;DR

The paper tackles how to help drivers contextualize driving sustainability by leveraging on-board CAN-bus data from ADAS/AV-equipped vehicles to produce a post-drive dashboard. It introduces a sustainability dashboard that computes and visualizes indicators for safety, fuel efficiency, and comfort, using metrics such as headway, TTC, a CAN-based fuel-consumption model, and acceleration-jerk based comfort. Key contributions include a defined data format for driver-indicator interaction, a low-latency CAN-signal pipeline, and a user-friendly interface with trend and comparison capabilities to foster trust and acceptance of autonomous driving assistance. The work demonstrates how exposing under-the-hood vehicle data in an interpretable format can support driver engagement, guide iterative controller development, and enable data-driven improvements in AV safety, economy, and ride quality.

Abstract

This work describes the use of on-board vehicle data from cars with advanced driver assistance features as a trip summary, with the goal of helping drivers contextualize their driving habits in terms of sustainability. The approach is similar to recent advancements in fitness tracking apps, which leverage smartwatches and other wearable devices to characterize activities during a workout or as part of daily fitness monitoring. Instead of adding new vehicle sensors, the data used for this work is from on-board driving data, namely, signals decoded from the vehicle's Controller Area Network (CAN) bus. With the deepening research of automatic driving technologies, Autonomous Vehicles (AVs) have gradually entered the consumer field, and more users are benefiting from the convenience and safety assistance provided by driving assistance and autonomous driving. However, various technical obstacles persist due to the complex environment, the non-communication of technologies, and users' trust. We propose indicators for evaluating the key characteristics of each drive, to facilitate drivers' familiarity with advanced driver assistance systems and to allow them to consider how different driving styles affect sustainability metrics. Further extensions will allow users to add feedback as part of the driving summary, laying a data foundation for future controller iterations based on real driving data and the attitude of drivers towards it.
Paper Structure (17 sections, 6 equations, 5 figures, 1 table)

This paper contains 17 sections, 6 equations, 5 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. Sustainability in Self-driving Development
  4. Visualization for Vehicle Performance
  5. Evaluations of AV Controllers
  6. Safety
  7. Fuel Efficiency
  8. Comfort
  9. Acquiring Data from AVs
  10. User-friendly Dashboard Design
  11. Overview
  12. Key Metrics
  13. Safety
  14. Fuel Efficiency
  15. Comfort
  16. ...and 2 more sections

Figures (5)

  • Figure 1: Overview of the proposed dashboard.
  • Figure 2: Safety analysis diagram showing the time headway values and speed values of categories of alert, attention and safe zones under conditions of cruise control ON and OFF.
  • Figure 3: Fuel efficiency diagram showing the efficiency plotted as cruise control state changes.
  • Figure 4: Comfort analysis diagram showing the speed, absolute acceleration, absolute jerk values of categories of comfort and discomfort under conditions of cruise control ON and OFF.
  • Figure 5: Trends of the safety index, fuel efficiency index, comfort index and cruise control ON percentage.