Transforming Experimental Education: Automation, Robotics, And Programming for Material Synthesis

Research Paper Recommended citation: Bawa, S. G., Ogunnoiki, A., & Chang, H. (2025). Transforming Experimental Education: Automation, Robotics, And Programming for Material Synthesis. In Kangaslampi, R., Langie, G., Järvinen, H.-M., & Nagy, B. (Eds.), SEFI 53rd Annual Conference. European Society for Engineering Education (SEFI), Tampere, Finland. DOI: 10.5281/zenodo.17631685. This Conference Paper is brought to you for open access by the 53rd Annual Conference of the European Society for Engineering Education (SEFI) at Tampere University in Tampere, Finland. This work is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 4.0 International License. TRANSFORMING EXPERIMENTAL EDUCATION: AUTOMATION, ROBOTICS, AND PROGRAMMING FOR MATERIAL SYNTHESIS A. Ogunnoiki a, H. Chang a, S. G. Bawa a, 1 a Department of Chemical Engineering, University College London, London, United Kingdom 1 Department of Chemical Engineering, University College London, London, United Kingdom, 0000-0001-5233-5476 Conference Key Areas: (1) Curriculum development and emerging curriculum models in engineering. (2) Digital tools and AI in engineering education. Keywords: Automation, Robotics, Education, Chemical Engineering, Materials ABSTRACT This study explores the integration of Opentrons OT-2 liquid handling robots into an educational framework, focusing on hands-on experimental learning and Pythonbased protocol scripting. Students from diverse academic backgrounds collaborated in interdisciplinary groups to leverage their respective strengths. Through the development of protocols using both Opentrons Protocol Designer and Python API, students gained valuable automation and programming skills. The transition from a graphical user interface to Python scripting emphasized efficient protocol design, reproducibility, and automation, equipping students with industry-relevant competencies. A key challenge identified was the limited accessibility to OT-2 robots, which impacted hands-on engagement. To address this, a visual simulation tool was developed, converting text-based Python simulation outputs into an interactive graphical representation. This tool enabled students to optimize pipetting strategies, minimize tip waste, and troubleshoot protocols in a virtual environment before executing them on the robot. The simulation platform also facilitated the integration of custom labware, supporting a wider range of experimental applications. The learning experience was further reinforced through gold nanoparticle synthesis using the Opentrons OT-2 robot, where students designed and executed experiments autonomously. The exercise highlighted the efficiency of automation compared to manual procedures. Additionally, students developed a deeper understanding of robot calibration, pipetting strategies, and hardware integration, enhancing their confidence in handling automated systems. This study underscores the importance of incorporating automation, programming, and simulation tools in laboratory education. By bridging the gap between computational and experimental skills, students are better prepared for data-driven scientific research and Industry 4.0 advancements. 1 Corresponding Author S. G. Bawa [email protected] 1 INTRODUCTION The rapid advancement of automation in Industry 4.0, driven by digital technologies such as Digital Twins, Artificial Intelligence, and Machine Learning, has sparked a new wave of research into human-computer collaboration. This has led to the emergence of Industry 5.0, which emphasizes the integration of human creativity, collaboration, and innovation alongside smart machines to design efficient and sustainable manufacturing processes (Mourtzis et al., 2022). A key vision of Industry 5.0 is placing human well-being at the center of manufacturing operations (Leng et al., 2022). As industry shifts towards a workforce equipped with Operator 5.0 skillsets, there is a growing need to prepare future chemical engineering graduates with more than just practical experience in robotics. Traditional problem-based learning (PBL) approaches in engineering education have been widely implemented worldwide, typically focusing on cognitive learning, interdisciplinary learning, and team-based learning (Kolmos et al., 2009). However, the evolving landscape of Industry 5.0 demands a more integrated approach to PBL—one that fosters creativity, interdisciplinary knowledge, and strong collaborative skills. In particular, project-based problem learning (PBPL) has proven to be highly effective in equipping students with industry-relevant competencies (Husin et al., 2025). Moreover, the integration of sustainability within PBL is gaining momentum in higher education (D’Escoffier et al., 2024) and is particularly crucial within engineering disciplines. Traditional laboratory formats in chemical engineering, such as material synthesis experiments, often rely on well-defined problems and step-by-step instructions provided by instructors. However, this structured approach may limit students' ability to think critically and innovate in problem-solving. To better prepare students for the demands of Industry 5.0, PBL should incorporate more open-ended challenges that encourage students to analyze real-world problems, explore the capabilities and limitations of available tools, and develop their own solutions. We hypothesize that allowing students to define and shape problems based on prior laboratory experience fosters a deeper understanding of automation technologies. When given access to advanced tools such as liquid-handling robots and model-building software, students can compare automated and manual processes, gaining insights into optimization strategies and the unique challenges introduced by automation. This iterative problemidentification and solution-development process is essential for cultivating the skills necessary for future chemical engineers. Opentrons OT-2 liquid handling robots are cost-effective robotic tools that had been widely used for research (Moufarrej et al., 2023) such as automation of biochemical assay (Moukarzel et al., 2024), with little attention of the use in for educating the next generation of scientists. More so the robotic platform had gain interest in the biological science (del Olmo Lianes et al., 2023) with little effort in field of material synthesis. Leverage on Opentrons OT2 inexpensive nature as compared to other robotic tools makes it an ideal candidate for education purposes within the sphere of material development and synthesis a major strength with the chemical engineering profession. This study aims to develop a structured curriculum for chemical engineering students with no prior experience in automated synthesis, focusing on optimizing material synthesis using a liquid-handling robot. A central component of this curriculum is simulation-based visualization, which plays a crucial role in troubleshooting and system optimization before deployment on physical systems. By implementing this curriculum, higher education students will acquire skills aligned with Operator 5.0, equipping them for the future of work in chemical engineering. 2 METHODOLOGY 2.1 Framework of the adopted procedure Students in the postgraduate taught program were divided into three groups comprising of six students in a group from both computational and experimental backgrounds. After a safety briefing, each group conducted a risk analysis and developed a risk assessment document. They then worked together to unbox, set up, and calibrate a liquid handling robot. The groups used Protocol Designer to create a protocol for writing the UCL logo on a 96-well plate, which was implemented on the robot (Fig.1a). Additionally, students developed a Python script to simulate the task, reducing the risk of robot damage (Fig.1b). These skills were later applied in gold nanoparticle (AuNP) synthesis. Fig. 1. Workflow for hands-on learning Liquid handling robot via (a) conventional route and (b) innovative approach 2.2 Preparation and calibrating the liquid handling robot To help students understand the components of a liquid handling robot and build their comfort with the system, they were tasked with assembling an Opentrons OT-2 robot. Students accessed resources like the physical manual or instructional videos from Opentrons’ YouTube channel, reflecting different learning styles such as visual, auditory, and kinaesthetic (Jayakumar et al., 2012). During the assembly, students collaborated to lift the 48 kg robot and removed protective brackets securing the pipette holder and gantry. This allowed them to explore the robot’s movement in 3D space and mount singleand 8-channel pipettes. These hands-on activities helped students gain confidence in handling the robot and deepen their understanding of robotic systems, which is crucial for engineering students. Liquid handling robotic systems, such as the Opentrons OT-2 and Flex, coordinate actions using X, Y, and Z coordinates along with a reference point, typically their starting position. Calibration is crucial for ensuring precise pipette movement in 3D space, accounting for variations in manufactured parts or movement in components like the OT-2 deck. Students gained hands-on experience calibrating both the OT-2 and Flex robots by performing deck, tip length, and pipette offset calibrations. This involved adjusting pipette tips through the user interface, moving the pipette along axes with different jump sizes (0.1, 1, and 10 mm). Completing this process emphasized the importance of reference points in automated systems, ensuring reliable and consistent results. It also strengthened students’ troubleshooting skills for calibration-related issues. 2.3 Custom logo exercise using protocol designer and implementation As part the PBPL methodology, after completing the calibration process, students were involved in a hands-on challenge to develop an efficient liquid-handling protocol capable of writing "UCL" logo on a 96-well plate with the aim to achieve the shortest possible run time. This involved transferring liquid from a reservoir into specific wells to form the letters U, C, and L. Students were introduced to Protocol Designer, a tool for developing protocols for Opentrons robots as illustrated in Fig. 2, allowing them to select the necessary pipettes and labware. The decision between single or 8-channel pipettes and 300 µL or 20 µL volumes emphasized efficiency. Using undefined labware could lead to errors, so students ensured accurate placement of hardware in the robot’s deck slots before executing the final protocol. Fig. 2. Labware definition and slot placement using Protocol designer 2.4 Custom Logo exercise using Python code and visualization While tools like Opentrons Protocol Designer simplify protocol development, scripting with the Python API offers additional educational benefits, such as programming skills and a deeper understanding of documentation. Unlike Protocol Designer, which requires explicit step-by-step instructions, Python allows for the use of loops, calculations, and conditional logic (e.g., if-else statements), enabling more flexible and efficient protocols. To reinforce these skills, students were tasked with recreating the same protocol they had designed in Protocol Designer using the Opentrons Python API. This process involved defining pipettes and labware, then translating transfer actions into Python code. This hands-on exercise introduced students with no prior coding experience to programming in an experimental context. The skills learned were later applied in an automated gold nanoparticle synthesis experiment. However, the limited access to OT-2 robots, with only three units available for six students, made it difficult for all students to test their protocols. To address this, the development of an online simulation platform for protocol testing was identified as a solution, offering a virtual environment for students to debug and refine their Python scripts, after which feedback was obtained from students. 3 RESULTS 3.1 Tables Table 1. Typical members of a group for the hands-on experimental lab session Student First degree discipline Background skill 1 Chemistry Experimental 2 Computer science and artificial intelligence Computational 3 Chemical engineering Experimental 4 Information technology and computing Computational 5 Computer science Computational 6 Machinery design and manufacture and its automation Experimental 3.2 Figures Fig. 3. Labware UCL logo protocol steps Fig. 4. Options in a ‘Transfer’ step using Protocol Designer Fig. 5. Reservoir used to hold the food dye, resulting UCL logo on 96 well plate and Python protocol for UCL logo Fig. 6. Visual simulation tool version 1, at two different points in a protocol run Fig.7. Image of developed visual simulation tool Fig. 8. Automated AuNPs synthesis using Python code by the students 4 DISCUSSION AND CONCLUSIONS Students with a computational background often excelled at coding tasks but struggled with hands-on experimental work. In contrast, students from a materials science background, who had primarily focused on material synthesis during their undergraduate studies, found the laboratory experiments more intuitive but faced challenges with programming. To foster collaboration (Bhat et al., 2020; Marra et al., 2016) and an inclusive learning environment (Korthals Altes et al., 2024), students were placed in diverse groups that balanced technical backgrounds and gender as presented in Table 1. The multidisciplinary nature of each group ensures that members could learn from each other’s strengths. The example group presented in Table 1 consisted of three males and three females, further promoting diversity and equitable participation in the learning process (Mills et al., 2011). After gaining confidence with liquid handling robot through calibration, students were introduced to various design parameters in Protocol Designer where they have to plan a strategy and iteratively refine their approaches in writing UCL logo. The most effective liquid transfer strategy developed by students is outlined in Fig. 3. An example of the inputs required in Protocol Designer to perform a transfer from the reservoir to a column in the well plate using the 8-channel pipette is shown in Fig. 4. The source labware and columns are selected, along with the destination labware. Tip handling settings define how tips are discarded—an important factor in optimizing waste and protocol efficiency. A poorly designed protocol would discard tips ‘Before every aspirate,’ whereas an efficient one would use ‘Per source well,’ allowing the same tip to be reused when aspirating from the same source liquid, provided mixing is not required. The final logo developed by the students can be seen in Fig. 5, alongside the reservoir used to hold the dye solution. Through this PBPL exercise, students were not only introduced to steps involved in protocol development but also critical design considerations in real world experiments such as time efficiency, precision in liquid handling, and sustainable lab practice. Similar PBPL approach extended to the rest of the module, and at the end, in a survey of students, all students reported they feel somewhat or extremely better prepared to use automation tools in industry/research after the module. In the industry, various liquid handling systems use different scripting languages. However, the experience of scripting in Python with Opentrons bridges the knowledge gap and equips students with transferable skills (Chadha, 2006) applicable to other environments. Additionally, teaching students to study API documentation and apply it independently encourages self-sufficiency when working with other systems. The Python API also allows for integration with external data and devices, a capability not available in Protocol Designer. Most importantly, it supports the simulation of protocols, outputting experiment steps in text format, which is essential for building the visual simulation tool discussed in this study. Fig. 5, shows the code developed by a student with comments (”#Description of what this line of code does”) for ease of understanding. Students hands-on configured the Temperature and Heater-Shaker Modules, adjusting temperature and RPM settings to explore their impact on experiments. The Temperature Module maintains a constant temperature, while the Heater-Shaker Module combines temperature control with adjustable shaking speed, both useful for isothermal material synthesis. Allowing students to develop and run their own protocols on the simulation tool improves their ability to create efficient protocols and troubleshoot effectively. There are 17 students in the postgraduate taught programme from which 85% reported that the simulation tool helps them troubleshoot or optimize protocols before using the physical robot. As they work with hazardous chemicals in more complex PBPL exercises like investigating reaction parameters for gold nanoparticle synthesis, testing and verification become crucial. An online simulation platform enables students to remotely test their protocols, ensuring steps are executed correctly. This approach is particularly useful for integrating custom labware, where understanding the robot’s interaction with new components is essential for preventing damage. By simulating such scenarios, students enjoyed benefit of PBPL exercises under safer and more protected environment According to the Mentimeter survey, 63% of students found the simulation tool easy to use, while 37% expressed a neutral stance. Notably, no students reported finding the tool difficult to use. Additionally, 76% of the students reported that the simulation tool helped reduce their anxiety and uncertainty regarding the use of the physical liquid handling robot. The Opentrons API allows for simulation through text-based output, but this can be difficult to interpret. To address this, a webbased simulation platform was developed using JavaScript, converting text outputs into an interactive animation that clearly shows labware positions, liquid levels, pipetting steps, and tip usage. The initial version of the simulator featured an interface to model the full Opentrons OT-2 deck layout, displaying well plates, reservoirs, and tip racks, with animated volume consumption from the reservoir upon aspiration, shown in Fig. 6. However, this version lacked the ability to simulate actions performed with an 8-channel pipette, as the Opentrons simulation text does not specify the instrument used for actions such as aspirate, dispense, pick up tip, and drop tip. Additionally, it was not able to simulate scripts involving custom labware. Opentrons provides a tool for creating custom labware, but its simulation tool does not support custom labware protocols. To address this, a Python script was developed to integrate custom labware definition files (.json) into the Opentrons Python library for simulation. This allowed students to create their own labware and test its viability using the visual simulation tool. Additionally, another Python script was created to improve the simulation output by including pipette actions in the run log. The final tool is modular, supporting the simulation of custom labware and 8-channel pipette actions as shown in Fig. 7. It tracks tip usage, labels labware slots, and enables color coding in reservoirs for easier identification of transferred liquids. A slider allows users to adjust simulation speed. To learn to integrate Python programming into robotics from PBPL approach, students were tasked with multifaceted experimental design problem involving gold nanoparticle synthesis, as shown in Fig. 8. Using the Turkevich method, they conducted 64 experiments to investigate optimal synthesis condition. They applied skills acquired in previous classes to program liquid transfer steps in Python, select the appropriate labware, and operate the heater-shaker module for heating and stirring. The educational value of PBPL in building students' experimental and programming skillsets is supported by survey results. On average, students reported prior experience levels of 3.3 out of 5 in Python programming and 2.0 in robotics. After completing the module, students reported 3.7 and an exceptional 4.5 in how much this module improve their confidence in Python programming and designing liquid handling protocols. These findings highlight how the PBPL approach effectively equips students with practical skills in programming and robotics, which are vital competencies for future data-driven optimization in Industry 5.0. This study highlighted the effectiveness of integrating Opentrons OT-2 robotics, Python scripting, and simulation tools in laboratory education. The simulation tool addressed accessibility challenges, enabling students to refine protocols and troubleshoot before physical implementation. The gold nanoparticle synthesis exercise reinforced automation's advantages, enhancing students’ confidence with robotic systems. This approach bridges computational and experimental skills, preparing students for data-driven research and automation industries. Future work will expand the simulation tool’s capabilities to support more hardware modules for teaching and research.