Take PyBaMM battery simulation from notebooks to team workflows
Ionworks gives battery R&D teams the data infrastructure and simulation management to turn open-source PyBaMM models into repeatable engineering processes.
Book a demo →
PyBaMM alone vs. PyBaMM with Ionworks
✕ Vendor-specific data parsing
✕ Parameters in notebooks and spreadsheets
✕ Manual simulation coordination
✕ No shared model versioning
✕ Results scattered across output folders
WITH IONWORKS
✓ Common format, all major cyclers
✓ Parameterized models with full provenance
✓ Managed simulation runs, automatic reuse
✓ Immutable, versioned, shared across the team
✓ Studies with linked, traceable results
Where notebook-based PyBaMM workflows break down
The challenge is not what PyBaMM can model. It's what happens when a team of five or fifteen people needs to run, compare, and trust those models week after week.
Battery parameter estimation is the real bottleneck
Physics-based battery models depend on accurate parameterization, and many parameters are unknown or difficult to measure directly. The main issue is identifiability: with current as input and voltage as output, there is very little information to determine individual parameter values. Each experiment is designed to amplify the contributions of a specific parameter subset while minimizing others.
Test data is fragmented across lab systems
Battery R&D teams generate data across BioLogic, Maccor, Neware, Arbin, and other cycler platforms. Each produces its own file formats with different column names, step conventions, and cycle definitions. Cell specifications, electrode compositions, and formation parameters live in spreadsheets updated on a different cadence than the cycling data they describe.
Scaling simulations past what notebooks can manage
Running a single simulation in a Jupyter notebook is straightforward. Running a structured study that compares dozens of protocol variations across multiple cell designs, tracks which parameterized model was used for each run, and lets a colleague reproduce results six months later requires infrastructure notebooks were never designed to provide. Teams end up building ad hoc systems of naming conventions, shared drives, and README files that break as soon as someone leaves or a project reorganizes..
Python flexibility becomes a coordination cost
When every engineer writes their own simulation scripts, each with different conventions for handling parameters, protocols, and results, the team accumulates coordination debt. A weekly simulation that three people run produces three slightly different workflows. A web interface extends access beyond the Python-fluent members of the team, but only if the underlying data and models are structured enough to support it..
How Ionworks extends PyBaMM for teams
Ionworks Studio is a battery modeling platform built on PyBaMM. It adds data management, parameterization tracking, and simulation coordination so models become shared, reproducible engineering processes.
01
Measure
Upload data from Maccor, Neware, BioLogic, Arbin, Novonix, and other cycler formats into one shared system of record. The ionworksdata Python library processes raw files through named readers and normalizes them into a common time-series format with automatically computed fields for capacity, energy, and step tracking.
Data is organized around cell specifications, cell instances, and cell measurements. Every measurement links to its cell and experimental context. Time-series data, step summaries, and cycle metrics are stored as connected layers. Teams script the full ingestion pipeline in Python, making it repeatable across projects.
02
Train
Build parameterized models through a visual modeling pipeline. Each parameterized model bundles a specific electrochemical model type (SPM, SPMe, DFN), a validated parameter set, and a cell specification into a single, ready-to-run configuration. Sensitivity analysis and identifiability checks help determine which parameters the experimental data can actually constrain.
Once used in a simulation, a parameterized model becomes immutable. To modify it, clone it, creating a new version with clear lineage back to the original. Every parameterized model links to the fitting process and ingested data that produced it. No more spreadsheets of parameter values with uncertain origins.
03
Predict
Select a parameterized model, define an experiment protocol, run the simulation. Protocols can be saved, typed as text, uploaded from cycler files, or built with the protocol builder. Studies group related simulations into focused investigations where teams run, compare, and analyze results in one workspace.
Identical completed simulations are reused automatically rather than rerun. The same parameterized model and protocol always produce the same result, and that result is stored once. Colleagues working from different studies that reference the same simulation see the same output.
04
Optimize
Sweep coating thickness, porosity, loading, and electrode geometry against real performance targets. Define objective functions tied to energy density, cycle life, charge time, or thermal headroom. Set constraints for voltage limits, temperature bounds, and plating thresholds.
The same parameterized models validated against test data now explore the design space. Results inherit the same reproducibility and traceability guarantees as individual simulations. A design sweep that would take weeks of physical testing completes in hours.
Example questions teams can answer
What is the fastest charge rate that avoids lithium plating across a temperature range?
A validated electrochemical model predicts plating onset across conditions that would take months to test exhaustively. Sweep C-rates and temperatures against plating thresholds in a single study.
How do we set charge and discharge limits for a new cell chemistry?
Parameterized models fitted to early test data let teams simulate operating envelopes before committing to long cycling studies. Define voltage, current, and temperature bounds from simulation rather than trial and error.
What are the energy-power tradeoffs for different electrode designs?
Sweeping design parameters in Ionworks Studio surfaces tradeoffs across the design space in hours. Vary coating thickness, porosity, and loading to map energy density against rate capability.
Which charge protocol minimizes degradation for a target duty cycle?
Protocol-driven simulations using real cycler files let teams compare protocol variants against the same validated model. Run degradation predictions across CC-CV, multi-step, and pulsed protocols.
Built by the team behind PyBaMM
Ionworks is built by the team that created and maintains PyBaMM. Ionworks Studio uses PyBaMM as its simulation engine, so teams get the same electrochemical model fidelity with the coordination and traceability their workflow requires.
Physics-based battery modeling is useful only when the models are parameterized against real data, the parameter sets are traceable, and the results are reproducible. Ionworks provides that as battery R&D software so teams don't have to assemble it internally.
Frequently asked questions
What is PyBaMM used for?
PyBaMM is an open-source Python library for physics-based battery modeling. Researchers use it to define, parameterize, and solve electrochemical models (SPM, SPMe, DFN, and others) that simulate battery behavior under different operating conditions.
Is PyBaMM enough for a battery R&D team?
For individual work, yes. For team-scale workflows, it lacks structured data management, versioned parameterized models, and shared access to results and simulations. Ionworks Studio adds those layers while keeping PyBaMM as the simulation engine.
What is a parameterized model?
It bundles an electrochemical model type, a validated set of parameters, and a cell specification into a single, ready-to-run configuration. Once used in a simulation, it becomes immutable. Clone to modify, maintaining clear lineage to the original.
Why use a GUI if my team already uses Python?
Python handles custom model development and scripted batch workflows. A shared web interface handles the routine work: configuring simulations, comparing results, reviewing parameterization fits, giving non-coding team members access. The Python SDK and REST API provide full programmatic access to the same data and workflows.
How does Ionworks relate to PyBaMM?
Ionworks is built by the team that created PyBaMM and uses it as the simulation engine. It adds structured data management, parameterization tracking, simulation coordination, and a web interface on top of the solver.
Does Ionworks replace my existing PyBaMM scripts?
No. Teams keep writing custom PyBaMM code when the problem calls for it. Ionworks handles the coordination layer: data ingestion and organization, parameterized model management, simulation runs and studies, and team-wide access to results.