The ARCOS Tool Portal: An Enabling Infrastructure for Hands-On Software Assurance Tools Exploration

ABSTRACT

Software certification for high-confidence systems increasingly depends on advanced
assurance tools to generate, analyze, and assess evidence of acceptable system risk.
However, these tools are often di?icult to access, evaluate, and adopt due to complex
installation requirements, specialized hardware dependencies, and steep learning curves.
As a result, researchers, students, practitioners, and certifiers face significant barriers to
exploring certification-relevant capabilities, comparing alternative approaches, and gaining
practical insight into the strengths and limitations of emerging technologies. The Automated
Rapid Certification of Software (ARCOS) Tool Portal addresses these challenges by providing
a centralized, web-based infrastructure that enables interactive exploration and
experimentation with a diverse set of software assurance and certification tools without the
burden of local setup or resource constraints.

The ARCOS portal serves as a centralized platform hosting simulators, verification tools, and
assurance tool suites for automated analysis of software assurance evidence. Designed as
a hands-on experimentation environment, users can "try out" tools directly through a
browser interface by launching tools and running curated example scenarios. By making
these capabilities accessible through a shared, community-facing infrastructure, the portal
enables users to explore how tools support system risk assessment, evidence generation,
curation, and assurance case construction. These scenarios are intentionally designed to
expose the kinds of artifacts, analyses, and outputs that matter in high-confidence software
contexts, such as evidence extraction, formal reasoning results, semantic relationships, and
assurance arguments.

The infrastructure also supports tool developers by providing a common deployment and
demonstration environment. Developers can integrate their tools into the portal using
standardized container environments, enabling rapid exposure to a broader user
community. This lowers the barrier for showcasing new capabilities, gathering feedback, and
understanding how tools are used in practice.

In this way, the portal functions as an interactive playground for high-confidence software
assurance and certification, supporting learning, evaluation, and early-stage adoption. The
infrastructure bridges the gap between research prototypes and practical use by making
sophisticated tools immediately accessible, lowering entry barriers, and fostering informed
exploration. This capability is particularly valuable for the high-confidence software systems
community, where trust, understanding, and usability are as critical as technical rigor.

In this presentation, we describe the design and implementation of the ARCOS Tool Portal,
highlight the range of software certification and assurance tools available through the
platform, and discuss our experience developing the underlying cloud-based, containerized
infrastructure that supports browser-based interaction. We also present the user-facing
tooling, curated example scenarios, and tutorials provided by the portal, illustrating how the 
infrastructure enables researchers and developers to explore, evaluate, and evolve software
assurance technologies in realistic, high-confidence contexts.

BIO

Dr. Yogesh Barve is a Research Scientist at the Institute for Software Integrated Systems (ISIS), Vanderbilt University, Nashville, TN. His research focuses on the intersection of model-driven engineering (MDE) and the development of performance-aware, safety-critical distributed systems spanning the cloud-to-edge continuum. Dr. Barve's work addresses the fundamental challenges of ensuring performance, reliability, and security across the cloud-to-edge continuum. Dr. Barve has extensive experience in the design of scalable, cloud-based data infrastructures for heterogeneous data, supporting high-stakes applications ranging from AI/ML-driven treatment selection in major depressive disorder to medical device cybersecurity. He has also contributed to the development of NIST's Universal CPS Environment for Federation (UCEF) Co-simulation Testbed Platform, advancing co-simulation frameworks for complex, federated autonomous systems. His work on interference-aware application performance modeling for multi-tenant systems (FECBench) was recognized with the Best Paper Award at the IEEE International Conference on Cloud Engineering. Dr. Barve holds a Ph.D. in Computer Science from Vanderbilt University, Nashville, TN, USA.