I’ve used and developed assays for both the Hamilton ML PREP and Opentrons Flex systems.
In terms of “out-of-the-box” usability, I would lean toward the ML PREP. However, you should first confirm that the hardware you want on deck is supported. The system can be somewhat limited in terms of available hardware, carriers, and the number of samples it can process in a single batch.
The software is very simple. When I used it, programming could only be done directly on the instrument itself (though I believe Hamilton has since added the ability to remotely access the system). Overall, it’s a great platform if you know exactly what you want to use it for—assuming the workflow fits within the hardware capabilities—and if you’re planning to run a small number of relatively fixed assays or processes.
This is especially true if you don’t have much experience with scripting or assay implementation. In many ways, the ML PREP simplifies automation by intentionally limiting what the system can do, which makes it very approachable. Because of these constraints, it’s also quite difficult to cause a hardware crash. The system checks the deck layout before running a method and walks the user through verifying that the correct labware is placed in the correct positions.
The Opentrons Flex, on the other hand, uses Python for scripting. If you’re familiar with Python—or willing to learn it—you should be able to work with the system. Opentrons provides a number of free resources for scripting, although it can sometimes be difficult to find very specific answers (for example, simple instructions for installing Python and setting up the Opentrons simulator).
The Flex offers significantly more flexibility because you can explicitly program each step of the workflow. However, that flexibility comes with trade-offs. It’s easier to make mistakes that could lead to hardware crashes—for example, entering an incorrect coordinate or placing labware incorrectly on the deck.
Another advantage of the Flex is that hardware can be added or removed relatively easily, making it much more versatile than the ML PREP in both software and hardware capabilities. That said, this also raises the barrier to entry.
In general:
If your team is comfortable with laboratory automation and willing to tinker and iterate, I would recommend the Opentrons Flex.
If you need a system that can reliably run a small number of processes with minimal setup and user intervention, the ML PREP is probably the better choice.
