2021 Research Review / DAY 2
Rapid Certifiable Trust
The DoD recognizes the need to field new cyber-physical systems (CPS) capabilities at an increasingly rapid pace, which is why it maintains a number of initiatives on rapid deployment. The demand for more rapid deployment, however, creates a need for verification techniques that can adapt to a faster deployment cadence, especially for CPS that are too big for traditional verification techniques and/or involve unpredictable aspects, such as machine learning.
The goal of Rapid Certifiable Trust is to reduce the deployment time of CPS by reducing the overall development and assurance times.
The goal of Rapid Certifiable Trust is to reduce the deployment time of CPS by reducing the overall development and assurance times. We will do this by enabling the use of unverified commodity software components (e.g., open source drone piloting software) guarded by verified enforcers that guarantee the containment of unsafe component behavior. We are developing compositional verification techniques to allow us to use multiple enforced components minimizing and automatically removing conflicting enforcer assumptions (e.g., reducing a plane’s airspeed to avoid a crash while increasing airspeed to prevent stalling). These techniques will allow us to assure full-scale systems, even if most of their functionality is implemented by unverified components.

In Context
This FY2020-22 project
- seeks to verify software-reliant systems that interact with physical processes (e.g., aircraft) to which existing verification technology does not scale
- will develop enforcing algorithms to identify unsafe control actions and replace them with safe actions
- drones are used to validate our approach in the SEI’s drone lab
- aligns with the CMU SEI technical objective to make software trustworthy in construction, correct in implementation, and resilient in the face of operation uncertainties
- also aligns with the CMU SEI technical objective to make software delivery timely so that the cadence of acquisition delivery and fielding is responsive to and anticipatory of the operation tempo of DoD warfighters
Principal Investigator
Dionisio de Niz
Technical Director, Assuring Cyber-Physical Systems
Principal Investigator
Bjorn Andersson
Principal Researcher
Anton Hristozov
Software Engineer
Mark Klein
Principal Technical Advisor
Bruce Krogh
Faculty Emeritus
Michael McCall
Associate Software Security Engineer
Gabriel Moreno
Senior Researcher
Amit Vasudevan
Senior Researcher
External Collaborators
Paul Griffioen
Phd Candidate,
Department Of Electrical And Computer Engineering
Carnegie Mellon University
John Lehoczky
Professor Of Statistics And Mathematics, Department Of Statistics & Data Science
Carnegie Mellon University
Raffaele Romagnoli
Postdoc Research Associate, Department Of Electrical And Computer Engineering
Carnegie Mellon University
Hyoseung Kim
Associate Professor, Department Of Electrical And Computer Engineering
University of California, Riverside
Ruben Martins
Systems Scientist, Computer Science Department
Carnegie Mellon University
Bruno Sinopoli
Department Chair And Das Family Distinguished Professor
Washington University In St. Louis