The initial conceptualisation followed this simple diagram below.
Central to this system is the OT2, which carries out the automation process. Its instructions are saved on a Jupyter notebook, which communicates with a main control computer with an interface for an individual to access. This main control computer also controls an xArm6 which, upon the completion of the automated assay, carries the microtiter assay plate to the spectrophotometer. The instructions to operate all the subsystems as previously described is stored in a GitHub repository.
Prior to any construction, a GitHub repository was made with the aim of storing all the programs that we have written for the subsystems. Physical construction of the subsystem didn’t take long as the OT2 was already available, and its connection simply required finding the right port to connect the OT2 to the main control computer. The calibration of the OT2, to an acceptable precision, took the course of approximately two weeks. Additionally, it took 2 weeks to set up connection to the computer via ethernet. Lastly, writing a series of programs instructing the movement of the xArm6 (inclusive of its gripper) took two weeks.
An initial difficulty with the OT2 was that it was struggling to pick up the pipette tips, as the OT2 started off by carrying the entire pipette tip box. We then contacted Dr Andrew Spiteri and Dr Daniel Winter, both of whom had recently operated on the OT2. They both suggested that the fault could be the tips, which they haven’t used in a while. Eventually, we realised that the tips were using were not the custom OT2 tips and as a result the entire pipette tip box was lifted from within its cell. However, in the process of arriving at that conclusion, we contacted the Argonne Laboratory, who provided us with more technical help on the internal structure of the OT2.
Another subsequent difficulty that we faced in the setup of the xArm6 was that the main control computer was not able to connect to the xArm6 via ethernet despite entering the correct IP address of the xArm6 control box. This obstruction eventually led us to contact the manufacturer of the xArm6 – UFACTORY. Upon contacting them, we realised that the control box is faulty. Consequently, it was shipped off to the manufacturer and fixed. This entire process took approximately 4 weeks to sort out.
Eventually, approximately 2-3 months after the start of the project, we were able to successfully run the commands on just the OT2 via Jupyter Notebook. Approximately one month after this, we were able to conduct a successful dry run with processes of the Carbonic Anhydrase Assay (without the input of any reagent/liquids). Subsequently, we were able to conduct a successful run mimicking the processes of both assays, where water replaced the bulk of our (liquid) reagents. Finally, we successfully ran a test urease assay with the true reagents. In both instances, the xArm6 was able to successfully pick up the microtiter plates and move it from the OT2.
Using the manual provided along the OT2 and its installed pipette, we first learnt how to turn on and off the OT2. Once this preliminary understanding is gained, we then proceeded to connect it to the main control computer, with which we calibrated the OT2. In terms of the associated software, we initially started with the standard Python script that was provided with the OT2 package. This script provided a program to conduct a serial dilution. Subsequently, we were able to manipulate the given script and develop our custom script to deal with both the urease assay and the carbonic anhydrase assay. Simultaneously to our development of software was its stepwise testing on the Jupyter notebook to ensure that each version of the software was valid, and errors did not aggregate.
Operating the xArm6 succeeded the OT2. This first involved stabilising and clamping it to a benchtop which required hands-on work. Using the given manual, we were able to connect it to the main control computer. Eventually, with the given interface of the xArm6 connected to the main control computer, we were able to devise how to manually alternate the position of each of its pivots – and ultimately write a series of commands on the interface that performs a series of pivot alternations as a collective command. Eventually, upon connecting the OT2 with the xArm6, the entire system was able to perform the automation of assays as well as move the microtiter plate to the spectrophotometer.