CAST

Connected Autonomous Safe Technologies

Texas A&M Engineering Experiment Station
  • Home
  • About Us
  • People
  • Research
    • Active Research Projects
    • Archived Research Projects
  • Publications
  • Contact Us

Formalizing Human-Machine Communication in the context of Autonomous Vehicles

Background

Tacit communication is often used by drivers and pedestrians in daily driving. Some examples include pedestrian crossings at traffic lights and stop signs, parking into a given spot when there are multiple cars waiting, reversing out of a parking lot or garage, getting out of a parked spot into traffic lanes, four way stop signs, etc. The need for such communication continues to exist when the driver is replaced by automation. The need becomes acute when plying in very dense traffic, such as autonomous golf-carts in airports or major events.

Without appropriate consideration for such communication between an autonomous vehicle and the pedestrians, the likelihood of accidents and missteps will be high. Consequently there is a need to understand and develop methodologies to incorporate such communication in autonomous driving decision making. This is the primary research objective of the proposed work.

Research Objectives, and anticipated Benefits

Our intent is to explore the “natural” human communication that happens when humans interact while in motion, and leverage that communication as much as possible in the autonomous decision making that will be replacing human decision making.

There are several important research questions that can be answered in this context, such as:

  1. What are the basic “language constructs” for such communication – that will be adequate for a very large set of driving scenarios, and can be precise in terms of syntax and semantics? While our focus will be on communication between the driver and the pedestrian, we will also peripherally explore tacit driver-driver communications that are commonplace during driving.
  2. What are the best mechanisms for autonomous-vehicle-to-pedestrian communication, and vice-versa? In particular, we would like to look at visual and aural communication mechanisms that would prove effective in such communication.

How can such communication be used to enhance both the contextual awareness and decision making capabilities in autonomous cars? Specifically could we detect the propensity for movement of a pedestrian, could we make optimal decisions on driving that leads to safer traffic scenarios.

Research Results

We are mounting and deploying programmable outdoor LED arrays and directional microphones and speakers on a Polaris GEM Golf Cart (the Dedicated Lane Low Speed Autonomous Vehicle TDP) as a prelude to performing initial experiments to understand pedestrian response to vehicle-initiated communication.

Project Status

Just starting.

TDP Alignment:

Dedicated Lane Low Speed Autonomous Vehicle TDP

Research Themes Alignment:

Driver / Autonomous Vehicle – Pedestrian Communication

Driver Behavior Characterization and Analysis

Researchers:

  • Dr. Swami Gopalswamy
  • Dr. Srikanth Saripalli
  • Dr. Sue Chrysler
  • Quang Le
  • Collaborators
    • Jeffrey Hickman (VTTI)
    • Scott Geller (VT)
    • Micha Roediger (VT)

 

External Links:

  • Safe-D Project Site

CAST
305 Doherty Bldg,
242 Spence St,
College Station, TX 77840

castinfo@tamu.edu

Texas A&M University College of Engineering
  • Accessibility
  • State Links and Policies
  • Privacy Policy
  • Website Feedback
  • Texas A&M Engineering Experiment Station

Copyright © 2022 · Texas A&M Engineering Experiment Station · All Rights Reserved