Orr Biologicals pioneers AI systems, computational models, and intelligent scientific instrumentation for biological research and discovery. We combine machine vision, physics-informed digital twins, and microfluidic observation to make complex living systems measurable, searchable, and safe to study.
Research concepts and computational tools in active development. Transparent methods; no validated hardware performance or commercial deployment claims.
THE SCIENTIFIC ARCHITECTURE
Living cultures drift, self-shade, divide, and adapt across timescales that pure black-box software cannot safely govern. At Orr Biologicals, artificial intelligence is never a substitute for biological measurement—it is an instrument for interpreting variation, testing counterfactuals, and proposing bounded experiments.
Bulk averages hide rare phenotypes and early contamination. By pairing microfluidic compartmentalization with time-series microscopy and computer vision, we design workflows that turn individual cellular trajectories into structured, queryable records.
Rather than fitting unconstrained neural networks to sparse sensor logs, our computational biology work embeds mechanistic equations—Droop nutrient quotas, Steele photoinhibition, and Beer–Lambert light transport—directly inside the forecasting loop.
In wet-lab and cultivation environments, an uncertain model proposal must never move a pump. Our instrumentation architecture separates AI analysis from physical actuation through independent software validators, firmware clamps, and physical stops.
RESEARCH PORTFOLIO · MULTI-SCALE SYSTEMS
Microalgae and cyanobacteria serve as our primary model organisms.
Both platforms are research concepts in development, not validated instruments.
AI-driven microalgal cultivation, multi-channel environmental sensing, physics-based growth modeling, and bounded control. Designed to make the chemical and optical conditions around a living culture legible and testable.
SENSE → MODEL → VERIFY → ACT Explore Algaephyte cultivation research ↗
An autonomous environmental microfluidic platform concept for isolating, imaging, analyzing, and investigating individual microalgal cells—converting biological diversity into searchable single-cell datasets before making phenotype claims.
ISOLATE → IMAGE → COMPARE → VALIDATE Explore Cyanoflow single-cell research ↗EVIDENCE & SCOPE
Scientific credibility requires separating what works in software today from hardware and biological hypotheses still awaiting experimental validation.
Interactive computational biology model
Adjust incident irradiance, dissolved inorganic carbon, and gas exchange. The mechanistic model recalculates growth rate, oxygen stress, and optical attenuation in real time.
Simulated response from a growth model — illustrative, not measured laboratory data.
This normalized teaching model combines a peaked light response, saturating carbon availability and an oxygen-stress penalty. Biomass assumes an 18 L vessel and 0.13 g/L/day reference productivity; protein and CO₂ use fixed conversion factors. The 72 h curve is a relative illustration, not a calibrated time-series prediction. Inputs do not account for species, temperature, nutrients, culture density or measurement uncertainty. Do not use these outputs as dosing or food-safety advice.
RESEARCH, LITERATURE & INTERACTIVE LABS
How we separate peer-reviewed literature, computational models, and unvalidated hardware concepts.
02 / RESEARCH JOURNALAI-assisted literature summaries alongside deep technical notes on digital twins, sensors, and microfluidics.
03 / DIGITAL WORKSPACESTest cultivation dynamics in the Algal Digital Lab or explore the interactive BioRealm research environment.
COMPUTATIONAL BIOLOGY · ALGAEPHYTEDigital twins for biological cultivationState variables, counterfactual testing, and why simulation-first beats actuate-first in living systems.
SINGLE-CELL DISCOVERY · CYANOFLOWSingle-cell analysis with microfluidicsHow droplet compartments, time-series imaging, and recovery turn cellular diversity into searchable data.Specialized domain hubs: Algae biotechnology · Microalgae cultivation · Photobioreactors · Scientific glossary · Algal Digital Lab · Interactive research lab · Expedition demo
WHY ORR BIOLOGICALS
We started Orr Biologicals because we were shocked by how hard it was to access algae biotechnology and modern biological instrumentation. An early interest in biology, followed by college coursework and microbiology, led us to build open computational tools, transparent simulations, and intelligent hardware concepts. We share our research in the open while developing the technology; our platforms are not available for purchase or deployment.
Proposed architecture
Six proposed layers, from the culture vessel to optional parameter sharing. This is an architecture to develop and test, not a report from deployed instruments.
Algaephyte · design notes
A column of spirulina, six sensor readings, a small model that forecasts growth, a camera that is not allowed to act, and an optional network that shares what vessels learn — never what they see.
The planned inputs are pH, temperature, wall irradiance, optical density at 750 nm, dissolved oxygen and conductivity. We want to use them to estimate biomass, internal nitrogen quota and dissolved inorganic carbon. Droop quota kinetics, Steele light response and Beer-Lambert attenuation are candidate model components; fitting them to real culture data is work still ahead.
The forecast target is seventy-two hours. A local rule-based planner would compare possible changes to carbon supply, nutrients, airflow and light. An optional assistant could explain proposals or accept operator requests, but would not have direct actuator authority. We need to test whether the model remains useful over that forecast horizon.
The proposed safety structure has three layers: fixed limits on mass, duty cycle, pH, temperature and dose frequency; a model check on the proposed action; and independent limits on the ESP32 actuator controller. A watchdog and physical emergency stop are also design requirements. We cannot call these protections reliable until fault testing demonstrates them.
Optical density measures light attenuation, not organism identity. A camera could provide a second evidence stream by looking for coils, fragments and unfamiliar shapes. We are considering a quantized vision model on a Coral accelerator, with training and compilation performed separately. A useful classifier needs labeled images and tests against bubbles, debris and organisms it has not seen before.
We have not established inference speed, power draw or contamination-detection accuracy for this proposed setup. Vision should be able to request caution, not authorize a pump. That separation is a design requirement, not an experimental result.
The proposed controller uses a Raspberry Pi 5 for telemetry, the growth model and local policy, with a Coral USB Accelerator for camera inference. A local Qwen assistant is an optional interface idea, not a dependency for control. Resource use, thermal behavior and safe fallback operation all need testing on the assembled hardware.
Algaephyte Mesh is a proposed way to share fitted model parameters rather than photos or raw traces. Candidate parameters include μ_max, I_opt, K_s, T_opt and Q_min. MQTT is the planned transport. Whether parameters transfer usefully between vessels and strains must be tested; we are not claiming that a network already exists or speeds up cultivation.
This educational calculator illustrates a mass deficit and an assumed staging rate of 0.5 g/L per hour. That rate is not a validated Algaephyte protocol or a recommendation for your culture. Verify your medium, measurements and procedure before any dosing.
Inputs are expressed as NaHCO₃ equivalents. Sodium carbonate and sodium bicarbonate are not interchangeable. A calculated deficit alone does not establish a safe dose.
The sensor stack
The proposed sensor suite measures water chemistry, light and optical density. No single reading can establish culture health, and the combined interpretation will need calibration and testing.
The digital twin
We plan to use nutrient-quota and light-response equations to evaluate proposed actions. Forecast accuracy, dosing limits and biological outcomes still need validation.
Droop's cell-quota kinetics decouple uptake from growth: cells store nitrogen, and division depends on that internal quota rather than the concentration in the medium. That single idea explains why a starved culture keeps dividing after you feed it, and why over-dosing nitrate buys you nothing but bacteria.
Steele's curve handles light — growth rises to an optimum irradiance and falls again under photoinhibition. Brightest is not best: past the optimum, the outer shell of the culture bleaches, oxygen supersaturates, and photosynthesis starts to poison itself. The twin treats irradiance as a control surface, not a schedule.
Beer-Lambert gives every radial shell of the vessel its own light climate, so the outer millimetre of a dense culture can be photoinhibited while the axis sits below compensation. A single wall sensor cannot see this. The twin can, because it integrates Steele's curve through the radial shells.
Most growers chase pH because pH is the number a ten-dollar probe will give you. In an alkaline Arthrospira medium the bicarbonate/carbonate pool is simultaneously the inorganic carbon supply, the overnight buffer, and the reason almost nothing else can live in the vessel. pH is what that pool looks like from the outside. The twin treats dissolved inorganic carbon as a state variable, not a setpoint.
As dissolved oxygen climbs through the afternoon, photosynthesis starts to supersaturate — the culture poisons itself. The first proposal is usually more air. The second is a modest dim. The third, if you have been stubborn about air, is a siesta. Oxygen stress is a state variable in the twin, not a footnote.
Autonomy with limits
An uncertain proposal should not move a pump. The design calls for three independent checks and a physical stop. Until tested, these are safety requirements, not guarantees.
A request such as "raise alkalinity to target" leaves out essential information: the mass, rate, current chemistry and equipment limits. Our design must reject incomplete requests rather than guess. We plan to test malformed proposals, stale sensor values, lost communications and stuck outputs before trusting automatic dosing.
Algaephyte Mesh
Our proposed mesh would share fitted growth-model parameters, not photos, raw sensor traces or locations. Participation would be optional. The privacy boundary and usefulness of shared parameters both need verification.
The transport design uses MQTT QoS 1, authenticated connections and TLS for remote traffic. Small payloads would suit limited connections. A last-will message can indicate a lost connection, but cannot establish whether a culture is healthy or a pump is working.
The proposed sharing concerns the growth model, not on-device camera training. We need to establish how to handle outliers, incompatible strains and malicious inputs. Local cultivation should not depend on the broker; disconnect tests will need to confirm that.
Local control by design
The proposed architecture separates the growth model, camera analysis and actuator controller. Our goal is local operation with independent hardware limits and a physical stop. Disconnect and fault testing must demonstrate that behavior before we claim it.
Inspect the hardware stackScale without fantasy
This number is arithmetic, not a result from our equipment: an assumed productivity of 0.13 g/L/day multiplied by a proposed 18 L volume gives 2.34 g/day. It does not demonstrate yield, food safety, carbon removal or scale-up performance. Those claims require measured runs, documented conditions and independent checks.
Development record
These are open design questions, not dated records of experiments we have completed.
When it goes wrong
Sensor drift, dry pumps, contamination and loss of power are risks the design needs to handle. These notes describe checks to investigate, not failures observed on deployed Algaephyte units.
Questions, answered plainly
Research correspondence · collaboration
Algaephyte and Cyanoflow are in development, not available for purchase or deployment. Tell us what biological questions you want to study, which instrumentation barriers you face, or how you could help test the designs. This form opens your email client; nothing is sent until you send the message.