Job Description-Digital Twin Lead

Role Snapshot

Position Title

Digital Twin Lead

Function

Product (Research, Innovation & Digital Engineering)

Reports To

Chief Research & Innovation Officer / Head - AI/ML & Digital Twin

Experience

812 years

Location

Bengaluru • WFO • Periodic site travel

Employment Type

Full-Time

Grade / Band

Role Summary

The Digital Twin Lead is a hands-on technical leader responsible for architecting, building, and scaling the next-generation digital twin platform for intelligent pump network infrastructure. The role blends physics-informed modelling, AI/ML, GPU-accelerated simulation, and immersive visualisation to create high-fidelity, real-time twins that simulate, predict, and optimise pumping systems across their operating envelope.

This is a builder's role, the successful candidate will code, model, and prototype alongside the team while setting technical direction, driving research-to-product translation, and shaping the twin as a core innovation asset.

Key Responsibilities

Simulation & Physics-Informed Modelling

- Design and build high-fidelity simulations of diverse pumping environments - centrifugal, positive-displacement, multistage, submersible, and networked pump systems across steady-state and transient regimes.
- Develop physics-informed neural networks (PINNs) and hybrid models that fuse first-principles hydraulics, thermodynamics, and structural mechanics with data-driven learning.
- Model complex phenomena: cavitation, water hammer, two-phase flow, pump-curve degradation, bearing wear, and network-level hydraulic transients.
- Couple reduced-order models with CFD/FEA where appropriate to balance fidelity and real-time performance.

Digital Twin Product Development

- Own the end-to-end twin product lifecycle architecture, data model, simulation core, ML layer, APIs, and immersive front-end.
- Build the twin as a living, self-calibrating system that ingests live telemetry (SCADA, IIoT, OPC-UA, MQTT) and continuously reconciles simulated vs. observed behaviour.
- Deliver operator, engineer, and executive-facing experiences using immersive technologies such as 3D visualisation, AR/VR/XR, and photoreal rendering.

NVIDIA & GPU-Accelerated Stack

- Lead adoption of the NVIDIA Omniverse / OpenUSD ecosystem for collaborative, real-time twin development.
- Leverage NVIDIA Modulus for physics-ML, CUDA / cuDNN / TensorRT for accelerated inference,



and Isaac Sim / PhysX where robotic or mechanical interaction modelling applies.
- Optimise simulation and ML workloads on GPU clusters (DGX / cloud GPU) for scale and latency targets.

AI/ML Engineering

- Architect ML pipelines for predictive maintenance, anomaly detection, energy optimisation, and prescriptive control.
- Apply modern techniques - graph neural networks for network topology, reinforcement learning for control strategy, surrogate models for real-time inference, and generative approaches for synthetic scenario generation.
- Establish MLOps, model governance, drift monitoring, and validation practices.

Leadership & Innovation

- Translate research prototypes into production-grade twin capabilities.
- Mentor a multidisciplinary team of engineers and researchers; foster a culture of rigor, curiosity, and hands-on experimentation.
- Partner with academic institutions, NVIDIA, and technology vendors on joint R&D.;
- Contribute to IP, publications, patents, and industry thought leadership.
- Represent iPUMPNET's twin capability to executives, customers, and at global forums.

Essential Qualifications & Experience

- Bachelor's or Master's in Mechanical, Electrical, Chemical, Aerospace, Computer Science, or Computational Engineering. PhD strongly preferred in simulation, CFD, physics-informed ML, or a related field.
- 8–12 years of hands-on experience building digital twin products, with demonstrable delivery of twins that combine simulation, AI/ML, and immersive technologies (AR/VR/XR, 3D real-time visualisation).
- Deep expertise simulating pumping and fluid infrastructure - pumps, pipelines, valves, manifolds, and networked systems across normal, transient, and fault conditions.
- Deep programming skills in Python and C++ (essential), with production-quality code, performance profiling, and software engineering discipline. R, MATLAB, or Julia for statistical, control, and calibration workloads (desirable).
- Solid hands-on experience with NVIDIA technologies - Omniverse, Modulus, CUDA, TensorRT, Isaac Sim, or equivalents.
- Proven expertise in physics-informed modelling and simulation — PINNs,



hybrid ML-physics models, surrogate modelling, or differentiable simulation.
- Strong AI/ML foundations - deep learning frameworks (PyTorch, TensorFlow, JAX), classical ML, and modern architectures (GNNs, transformers, RL).
- Working knowledge of hydraulic solvers applied to pump-network modelling.
- Experience integrating twins with OT/IIoT data sources - OPC-UA, MQTT, Kafka, and industrial settings.

Desirable

- Prior work on water networks, oil & gas, process industries, district energy, or desalination.
- Contributions to open-source simulation, physics-ML, or twin frameworks (e.g. WNTR, Modulus, OpenFOAM, SU2).
- Experience with FMI / FMU-based co-simulation and 1D/2D/3D system-simulation tools such as Ansys Twin Builder, Simcenter Amesim, Modelica / Dymola, MATLAB / Simulink for pump-motor-VFD trains.
- Familiarity with OpenUSD, Unreal Engine, or Unity for immersive twin front-ends.
- Publications, patents, or conference talks in digital twin, PINN, or GPU-accelerated simulation.
- Awareness of IEC 62443 (OT cyber), ISO 23247 (digital twin), ISO 55000 (asset management), and regional cybersecurity frameworks (NIS2 in EU, CEA guidelines in India).

Key Competencies

Hands-on technical depth • Research-to-product mindset • Systems thinking across physics, data, and software • GPU-performance intuition • Ability to lead and code • Strong communication with both scientists and operators • Bias for prototyping and rapid iteration • Intellectual rigor.

Success Measures (First 12–18 Months)

- Reference twin architecture built on GPU-accelerated, physics-informed foundations, adopted as the iPUMPNET standard.
- Live, calibrated twins deployed for priority pump assets/networks with validated accuracy (3–5% on head, flow, and power).
- At least one physics-informed ML capability (e.g., cavitation prediction, transient forecasting, energy optimisation) in production use.
- Immersive twin experience (Omniverse / XR-based) demonstrated to executives, customers, and operations teams.
- High-performing, multidisciplinary team established; research pipeline feeding product roadmap.
- Quantified outcomes benchmarked against pre-twin baseline: 10% energy consumption reduction, 20% reduction in unplanned pump downtime, and measurable non-revenue-water reduction on pilot deployments (aligned with industry benchmarks from Sulzer, Xylem MAS 801, and Bentley WaterSight case studies).

📌 Head of Digital Twin -Utilities or Pumping Stations (Bengaluru)
🏢 Pump Academy
📍 Bengaluru