A TEM experiment is not just a sequence of images, it is a scientific event.

‍

During an in-situ experiment, the sample may be heated, cooled, biased, exposed to gas, placed in liquid, mechanically loaded, tilted, moved, illuminated, or monitored over time. Each of those conditions changes the meaning of the image.

‍

If the metadata is missing, fragmented, or trapped inside separate software systems, the experiment becomes harder to repeat, compare, search, analyze, and reuse.

‍

Automation depends on context such as:

‍

• Holder state
• Temperature
• Voltage
• Current
• Resistance
• Gas flow
• Pressure
• Liquid flow
• Electrochemical conditions
• Motion or position
• Controller status
• Microscope conditions
• Detector or camera timing
• Session notes
• User-defined events
• Sample identity
• Time stamps
• Safety limits and alarms

‍

Without this information, automation becomes shallow. The system may move, acquire, or repeat a command, but it does not understand the experiment. Colibri helps provide the experimental context that automation needs.

AI-ready microscopy requires more than pixels.

‍

A model cannot learn from information that was never recorded. A future researcher cannot reproduce conditions that were never captured. A shared facility cannot mine historical data if the data lacks searchable metadata. An autonomous workflow cannot make reliable decisions if the system cannot connect image changes to experimental conditions.

‍

AI-ready TEM data should help answer:

‍

• What was the sample?
• What microscope conditions were used?
• What holder or sample-environment system was active?
• What changed during the experiment?
• When did the change occur?
• Which image, spectrum, or signal corresponds to that change?
• What did the user observe?
• What data was exported?
• What tools can read the result?
• Can the experiment be searched, compared, reproduced, or reused later?

This is why metadata matters.

‍

Colibri supports AI-ready microscopy by helping preserve the experimental record in a structured and portable way. It does not claim that one software platform should own all future AI analysis. Instead, Colibri helps prepare the data so users, OEMs, open-source communities, national labs, and institutional data systems can build on top of it.

‍

The first step toward AI-guided TEM is metadata-ready TEM.

Colibri’s role is intentionally focused. It is designed to help collect, synchronize, and export the experimental context created by Hummingbird in-situ systems and related microscope workflows.

‍

Colibri can support:

‍

• Holder and controller metadata capture
• Timestamped experiment records
• Synchronization between sample-environment data and image timing
• Basic experiment logging
• User notes and event marking
• Control-state history
• Safety and alarm records
• Structured export for downstream use
• Open handoff to OEM, institutional, third-party, and open-source workflows

Colibri is not intended to become the customer’s only software environment. It is not intended to replace the microscope control interface. It is not intended to replace established analysis tools. It is not intended to force researchers into a single vendor-controlled workflow.

‍

Colibri’s value is in preserving the in-situ experimental context and making that context usable outside the original acquisition session.

‍

That is the foundation automation needs.

‍

A metadata-ready experiment record creates a stronger foundation for future automation.

‍

OEMs, facilities, researchers, and software developers can build workflows that use structured Colibri data as an input. These workflows may include microscope automation, data review, image analysis, sample tracking, facility reporting, reproducibility checks, experiment planning, or AI-assisted decision making.

‍

Colibri does not need to own those layers.

‍

Instead, it helps make them possible by providing cleaner experimental context. For microscope OEMs, this means Colibri can complement the microscope operating environment without trying to replace it. For users, this means the data can move into the tools they already trust. For facilities, this means in-situ experiments can become easier to document, search, compare, and archive. For future AI workflows, this means image data can be connected to the experimental conditions that produced it.

‍

Good automation does not require every tool to come from one vendor. It requires the right information to be available at the right time in a usable format.

‍

The microscopy community already has powerful software ecosystems; researchers use OEM tools, Python scripts, Jupyter notebooks, HyperSpy, py4DSTEM, LiberTEM, institutional databases, custom lab pipelines, cloud storage, and high-performance computing environments. New tools will continue to emerge as microscopy, AI, and data science evolve.

‍

A closed software platform cannot predict or own all of that, which is why Colibri is designed around open handoff. The goal is to help produce structured records that can be used beyond Hummingbird software. That may include CSV exports, JSON records, HDF5 or Zarr-style data structures, Python-readable metadata, facility databases, or future formats required by customers and collaborators.

‍

This gives researchers flexibility to collect in-situ experimental context with Hummingbird systems, then move that data into the workflow that makes sense for their lab.

‍

Open-source tools are not a threat to Colibri, they are part of the reason Colibri exists.

Colibri is deliberately not positioned as a closed end-to-end TEM software suite.

‍

It does not try to replace:

‍

• Microscope OEM software
• Camera acquisition software
• Facility LIMS systems
• Institutional storage systems
• Open-source analysis tools
• Python workflows
• Cloud or HPC pipelines
• Research-group-specific data practices

‍

Colibri’s role is narrower and more practical: It preserves the in-situ experimental context and makes that context portable.

‍

That means researchers can use Colibri data with the software and data systems that already work for them.