Date of Award


Document Type


Degree Name

Master of Science in Computer Engineering


Department of Electrical and Computer Engineering

First Advisor

Scott R. Graham, PhD.


Widespread use of fully autonomous vehicles is near. However, the desire for a human to maintain control, even if limited, of a vehicle will likely never fully subside. Protocols to safely and efficiently manage reservation-based intersections with a mixture of fully autonomous, semi-autonomous, and non-autonomous vehicles exist such as AIM, SemiAIM, and H-AIM. Missing from these protocols is persistent human control of semi-autonomous vehicles in approaching and navigating autonomous intersections without the use of traditional signals. This thesis offers a proof-of-concept of a reservation-based protocol with necessary extensions required for human control in semi-autonomous vehicles. Desired is a protocol that maintains the benefits in efficiency of a fully autonomous environment, such as AIM, while allowing persistent human control of a vehicle. Proposed are possible feedback mechanisms for human response such as displays detailing intersection arrival time, goal velocity, lane keeping assistance, and other warnings. Also developed is a synthetic environment able to demonstrate cyber attacks, their mitigations, and aid in designing a protocol introducing persistent human control. The AFTR Burner three-dimensional virtual world offers the ability to model this physics based environment in a highly predictable and realistic manner. The reservation-based protocol used in the synthetic environment is first verified and validated against both an established reservation-based protocol, such as AIM, and also use case scenarios to determine if the expected behavior is exhibited. Preliminary observations suggest that persistent human control is a possibility among reservation-based autonomous intersections, but further research must be done to determine its viability.

AFIT Designator


DTIC Accession Number