STREAM (Smart Tracking using Raman & Enhanced Analytics Monitoring) is an initiative to advance monitoring and control in continuous fermentation. It integrates real‑time analytics to track substrate use, biomass growth and metabolite formation to steer these processes. This improves efficiency, stability and scalability in both continuous biomass and precision fermentation. STREAM provides a data‑driven foundation for next‑generation fermentation with enhanced control and predictability.

The widespread adoption of biomass and precision fermentation for a wide variety of ingredients such as microbial protein, oils and/or sweeteners is hampered by high costs and limited predictability at scale. Current analyses indicate that a substantial share of total costs stems from capital charges (CAPEX >30%) and raw materials, with the carbon source contributing 40–50%. Further reducing raw‑material costs is difficult because yields are often close to maximum, while lower‑cost carbon sources come with higher contamination loads and complex polysaccharide structures that require resource intensive pre-treatments. Therefore in this initiative we try to approach overall cost-reduction by reducing the CAPEX. This reduces total costs, but importantly also upfront risk for investors.

The main challenge with the often used batch or fed-batch production is that these suffer from lost volumetric productivity during the lag phase and sometimes stationary phase. This loss can be minimized by moving to fed‑batch or (semi‑)continuous processes (potentially up to ~50%). However this comes with its own specific challenges as contamination or sub-optimal fermentation conditions may both lead to lower product yields on the carbon source. In this project we address this by exploring which process monitoring tools allow for tracking the relevant process parameters. This will allow ensuring that the feed for the fermentation is controlled such that the process always runs within an optimal parameter window.

PPP STREAM will deliver smart technologies and digital innovation via an advanced PLC tool hat predicts the processes required to obtain the most optimal fermentation conditions for the maximum length of time feasible for a specific fermentation process. In specific terms, PPP STREAMS will deliver:
- Selection and development of Process Analytical Technologies (PAT) for process monitoring
- Development of an MPC-based reinforcement learning controller to achieve optimal process control that takes into account batch-to-batch variation of the feedstock
- Development of a fed-batch or continuous fermentation process using a unique contamination control strategy by use of nutrient limitations
- Showcasing reproducibility of process monitoring and control of the extended fermentations at scale at the pilot/industrial facilities of the industrial partners
-Full Techno-Economic Model of the standard fermentation process and more continuous process designs. A generic version of this model is to become publicly available