eDICEML

Oak Ridge National Laboratory has developed eDICEML, a software-based digital twin that emulates the hardware and software of instrument-computing ecosystems. Designed to reduce cost and risk during development and explore design options, this platform enables continuous testing, orchestration, and deployment of complex scientific workflows without requiring access to physical instruments or infrastructure.

Description

eDICEML emulates scientific instrument-computing ecosystems (ICE) by replicating networked instruments, computing hosts, and interconnections using mininet software on Linux. It runs in two modes: natively on ICE systems (in situ) or as a portable virtual machine. This environment enables the execution of control and orchestration software identical to the physical infrastructure, facilitating seamless development and deployment. To improve the accuracy of network emulation, eDICEML integrates a custom machine learning method that adjusts virtual network measurements to closely reflect physical network performance. Instrument control software—whether vendor-specific or open-source—is executed on virtual hosts to simulate real workflows. This approach eliminates disruption to physical instruments, supports remote development, and accelerates production readiness for advanced software modules. Applications include scanning transmission electron microscopes and EPICS-controlled physics instruments.

Benefits

• Eliminates need for access to physical instruments during design and development
• Reduces cost, downtime, and risk of damage to sensitive equipment
• Enables exploration of infrastructure designs without building them
• Enables seamless transition of software from development to deployment
• Corrects network performance emulation using integrated machine learning

Applications and Industries

• Development and testing of software for scientific instrument operation and control
• Digital twins for research infrastructure in national laboratories and industry
• Design space exploration tool for networked instrument-computing ecosystems
• Network and software performance analysis for remote experimentation

Contact

To learn more about this technology, email partnerships@ornl.gov or call 865-574-1051.