Question
Answer
State and actions are provided with predefined template and require customization from automation tool users, right?
State and actions are provided with predefined template and require customization from automation tool users, right?
Yes. Currently, users define the states manually in an XML file, which includes rules describing how to detect each state (e.g., what should be visible on screen). You can also rely on AI to detect states, but it’s not yet 100% reliable. Once states are defined, users specify the actions for each state (e.g., “insert card,” “enter PIN,” “press cancel”). A UI-based editor is coming soon to simplify this process, but for now, it requires customization from the automation tool users.
Question
Answer
If a test fails halfway through, can the automation recover or continue from the next test case?
If a test fails halfway through, can the automation recover or continue from the next test case?
Yes, there are routines and recovery options built in. For example, if the terminal is asleep, a “setup routine” can wake it automatically. If a test fails, the recovery logic can return the terminal to an idle state by pressing specific key sequences (like “cancel” or “stop”). This allows the system to automatically reset and continue with the next test case.
Question
Answer
We’re currently using your solution for regression and stress testing. How can we upgrade it to use it for EMV Level 3?
We’re currently using your solution for regression and stress testing. How can we upgrade it to use it for EMV Level 3?
It’s straightforward.
You’ll need to add the PaytestHub EMV L3 Player add-on to your existing setup. If you use ICCSimTMat, you’ll connect via PaytestProbes; if you use Fime BTT, you’ll connect via SmartLink SmartWave or SmartConnect through the PaytestMux.
You’ll then configure your state and action files, and you’re ready to automate Level 3 testing.
Question
Answer
What needs to be prepared for automation to execute - from the user’s perspective?
What needs to be prepared for automation to execute - from the user’s perspective?
- Set up the agent locally and configure extensions (e.g., PaytestRobot, camera, ADB, ICC, or BTT).
- Connect the EMV L3 tool - either Fime BTT (with integrated automation module) or ICCSimTMat (with auto API).
- Prepare state and action files, ideally with support from PaytestLab.
- Load your test project (TSE file) in the L3 tool.
- Finally, run your test suite - automation executes all cases sequentially while you monitor results later.
Question
Answer
How long does a full L3 automated run take compared to manual testing?
How long does a full L3 automated run take compared to manual testing?
A full Visa contact and contactless test suite takes around 1.5 hours with automation. Manual execution may take longer due to breaks and human input. The biggest advantage isn’t just speed - it’s consistency and repeatability. Automated runs can execute unattended and deliver identical results every time, even during regression retests.
Question
Answer
Does it only work with Android devices or also with legacy terminals?
Does it only work with Android devices or also with legacy terminals?
It works with both.
Android terminals are common now, but legacy terminals are supported using OCR (camera-based) observation. You lose ADB integration, but automation still functions the same way through visual recognition.
Question
Answer
Will you integrate with other providers apart from ICC and Fime BTT?
Will you integrate with other providers apart from ICC and Fime BTT?
Yes - if other L3 tools expose APIs for automation.
Currently, only ICCSimTMat and Fime BTT offer this, but PaytestLab is open to integrating with tools like Payhuddle or Galitt if they release automation interfaces. Work is also underway to integrate PaytestProbes directly into third-party tools.
Question
Answer
Can both tools (ICCSimTMat and Fime BTT) be used simultaneously on one PaytestRobot?
Can both tools (ICCSimTMat and Fime BTT) be used simultaneously on one PaytestRobot?
Yes, with some setup considerations.
- If you use two terminals, one per tool, on a PaytestRobot T4, they can run in parallel on the same PC.
- If using one terminal for both tool, you’ll need a multiplexer between the probe and terminal for Fime BTT Smart Connect. The signal from the Smart Connect is then shared on the PaytestProbe.
- ICCSimTMat works directly with PaytestProbes “out of the box.”
Question
Answer
How do you handle ICC and BTT updates in the automation scope?
How do you handle ICC and BTT updates in the automation scope?
Currently, updates are manual.
When either tool releases new TSE or application files, these must be manually updated on the local PC. Similarly, exporting results (e.g., XML or receipts) still requires local access. Automating these update and export tasks is planned for a future release.
Question
Answer
Is it integrated with the Mastercard MAS Simulator?
Is it integrated with the Mastercard MAS Simulator?
That depends on the L3 tool.
If your certification tool (e.g., ICCSimTMat or BTT) integrates with the Mastercard MAS simulator, PaytestHub automation will support it transparently. This allows host logs to be verified automatically during testing, reducing “inconclusive” outcomes.
Question
Answer
In CI/CD, can we trigger a regression job that reuses the state and action framework but consumes a regression parameter matrix?
In CI/CD, can we trigger a regression job that reuses the state and action framework but consumes a regression parameter matrix?
Not yet - but it’s on the roadmap.
The goal is to allow regression runs to use a lightweight JSON input (rather than EMV test scripts) and reuse the state/action logic to perform automated regression testing. This feature is actively being developed.
Question
Answer
Are there any limitations you’re still working to overcome?
Are there any limitations you’re still working to overcome?
Yes, a few key ones:
- Manual validations - some test cases ask contextual questions (e.g., “confirm Visa Debit or Credit”) that require user verification.
- State/Action UI - currently managed via XML; a graphical interface is under development.
- Result verification — some host log checks still require manual review.
- API limitations - L3 tool integrations don’t yet support automated export or submission.