: Wastewater treatment using aerobic granular sludge has gained increasing interest due to its advantages compared to conventional activated sludge. The technology allows simultaneous removal of organic carbon, nitrogen, and phosphorus in a single reactor system and is independent of space-intensive settling tanks. However, due to the microscale, an analysis of processes and microbial population along the radius of granules is challenging. Here, we introduce a model system for aerobic granular sludge on a small scale by using a machine-assisted microfluidic cultivation platform. With an implemented logic module that controls solenoid valves, we realized alternating oxic hunger and anoxic feeding phases for the biofilms growing within. Sampling during ongoing anoxic cultivation directly from the cultivation channel was achieved with a robotic sampling device. Analysis of the biofilms was conducted using optical coherence tomography, fluorescence in situ hybridization, and amplicon sequencing. Using this setup, it was possible to significantly enrich the percentage of polyphosphate-accumulating organisms (PAO) belonging to the family Rhodocyclaceae in the community compared to the starting inoculum. With the aid of this miniature model system, it is now possible to investigate the influence of a multitude of process parameters in a highly parallel way to understand and efficiently optimize aerobic granular sludge-based wastewater treatment systems.Key points• Development of a microfluidic model to study EBPR.• Feast-famine regime enriches polyphosphate-accumulating organisms (PAOs).• Microfluidics replace sequencing batch reactors for aerobic granular sludge research.

Long‐term stable operation of bioelectrochemical systems (BES) presupposes the avoidance of mass transfer limitations of the electroactive biofilm. Excessive pH‐gradients from bulk to electrode interface or substrate limitations of the electroactive biofilm are known to diminish the electrical performance of BES. In this study the impact of the morphology of a mixed‐species electroactive biofilm cultivated on the electrical performance of a microbial electrolysis cell (MEC) was investigated to identify the optimal biofilm for real‐life applications in wastewater treatment. Noninvasive monitoring by means of optical coherence tomography and an industrial biofilm sensor allowed for a real‐time evaluation of the morphology of the biofilm. The maximum current density of approximately 3.5 A/m² was found for a mean biofilm thickness in the range of 100–150 µm, beyond which thicker biofilms caused mass transfer limitations. Along with local biofilm detachment a continuous decline in efficiency demonstrates the need for active biofilm control to adjust the biofilm thickness.

The paper introduces a fully automated cultivation and monitoring tool to study biofilm development in replicate experiments operated in parallel. To gain a fundamental understanding of the relation between cultivation conditions and biofilm characteristics (e.g., structural, mechanical) a monitoring setup allowing for the standardization of methods is required. Optical coherence tomography (OCT) is an imaging modality ideal for biofilms since it allows for the monitoring of structure in real time. By integrating an OCT device into the open-source robotic platform EvoBot, a fully automated monitoring platform for investigating biofilm development in several flow cells at once was realized. Different positioning scenarios were tested and revealed that the positioning accuracy is within the optical resolution of the OCT. On that account, a reliable and accurate monitoring of biofilm development by means of OCT has become possible. With this robotic platform, reproducible biofilm experiments including a statistical analysis are achievable with only a small investment of operator time. Furthermore, a number of structural parameters calculated within this study confirmed the necessity to perform replicate biofilm cultivations.