A Community Roadmap to Robust Science
in High-throughput Applications
2021 Virtual World Cafés


The RobustScience project gathers interdisciplinary communities to define a roadmap to robust science using high-throughput applications for scientific discovery. High-throughput applications combine multiple tasks into increasingly complex workflows on heterogeneous systems. Robust science should assure performance scalability in space and time; trust in technology, people, and infrastructures; and reproducibility or confirmable research in high-throughput applications.

We will define the roadmap through three Virtual World Cafés tailored to three groups: application users and scientists; system developers and users; and educators and trainers.

Wednesday, February 17, 2021
12-3pm EST, 9-noon PST
Wednesday, April 21, 2021
12-3pm EST, 9-noon PST
Wednesday, August 25, 2021
12-3pm EDT, 9-noon PDT

Project Mission

Three key concepts: High-Throughput Applications, Robust Science, and Roadmap

High-throughput applications are all those applications whose workflows are composed of independent operations (including computations and data analysis) and whose data processing is divided into independent transactions, at large and distributed scales.

Robust science assures performance scalability in space and time (scalability); trust in technology, people, and infrastructures (trustworthiness); and reproducible or confirmable research (reproducibility).

A roadmap defines future hardware architecture and software systems (e.g., tools, interfaces, libraries, data and model commons); programming models and compilers; algorithms and theory; principles and practices; workforce development and diversity.

One mission: Define a Roadmap

Our mission is to define a roadmap for establishing a vibrant, multidisciplinary community that works together to design, implement, and use a set of solutions for robust science applied to high throughput applications.

The roadmap will span across critical hardware and software areas (i.e., architecture, systems, HPC, programming models and compilers, and cybersecurity) combining these areas into a continuum through AI-orchestrations, policies, practices, and a diverse workforce.

Project Presentation