Biohybrid Devices Limited fuses cerebral organoids with micropatterned neurons into assembloids with predetermined connectivity. Two-photon fabrication defines the wiring, microfluidics keeps it alive, and a closed-loop interface lets your team run experiments on it — as a service, from your own lab.
Illustrative platform specifications shown for demonstration.
Assembloid · build BD-3Nlive
Running cohorts with research groups in neuroscience, biomedical engineering & pharma
Halden Institute for Neural SystemsMeridian NeurotherapeuticsAshgrove University · Biomedical EngineeringCambrian Autonomy LabsKestrel Bio FoundryNorthfleet Defence FuturesHalden Institute for Neural SystemsMeridian NeurotherapeuticsAshgrove University · Biomedical EngineeringCambrian Autonomy LabsKestrel Bio FoundryNorthfleet Defence Futures
Placeholder names shown for illustration.
01The platform
Four controlled stages between a cell line and a programmable circuit.
Most living-neuron platforms hand you a culture and hope structure emerges. We treat connectivity as a manufacturing specification — written, fused and verified before an experiment ever runs.
01
Derive
Cerebral organoids from your cell line
We culture region-specific organoids from iPSC lines — patient-derived, isogenic control, or our own reference bank — and QC every batch for morphology, marker expression and spontaneous activity.
iPSC · organoid
02
Pattern
Two-photon fabricated guidance scaffolds
Two-photon polymerisation writes 3D microchannel scaffolds at sub-micron resolution, while micropatterned neuron layers constrain where axons may grow. The wiring diagram is a design input, not an accident of culture.
2-photon · micropattern
03
Fuse
Microfluidic assembly into assembloids
Perfusion chips hold each organoid node in its own nutrient compartment and fuse them through the patterned tracts — so you get a defined multi-node circuit with long-horizon viability instead of a drifting blob.
microfluidics · perfusion
04
Interface
Closed-loop stimulation & readout
High-density electrode arrays plus optical readout close the loop between your protocol and the circuit. Send stimulus patterns, stream spikes back, and version every experiment against the exact connectivity map it ran on.
closed-loop · API
02Capabilities
Everything you need to run a living circuit like an instrument.
Built for neuroscientists, biomedical engineers and pharma teams who need reproducibility, not novelty.
Predetermined connectivity
Specify the graph — node count, tract topology, excitatory/inhibitory ratio — and receive an assembloid built to it, with a connectivity map delivered alongside every batch.
Two-photon 3D fabrication
Sub-micron polymerised scaffolds route axons through defined channels in three dimensions, so circuits are structured volumes rather than flat monolayers on a dish.
Microfluidic life support
Compartmentalised perfusion, oxygenation and waste clearance hold assembloids stable across long protocols — chronic dosing studies instead of single-session snapshots.
Closed-loop biocomputing
Millisecond-scale stimulate-and-read loops let you train circuits on control tasks, benchmark plasticity, and export the full stimulus/response record for analysis.
CNS disease models
Patient-derived and isogenic-control assembloids for epilepsy, neurodegeneration and neurodevelopmental phenotypes — with circuit-level endpoints, not just cell viability.
Research SDK & data
A documented API and Python client for protocol design, run scheduling and structured export, so wetware experiments version like the rest of your computational stack.
03Competitive positioning
Living neurons are not new. Designing their wiring is.
Here is how our platform sits against the best-known names in the field — and the gaps we built the company to close.
Comparison of Biohybrid Devices against Cortical Labs, FinalSpark and Koniku
Dimension
Biohybrid Devices
Cortical Labs
FinalSpark
Koniku
Neural substrate
3D assembloids — organoid nodes fused through micropatterned axon tracts
Cortical neuron cultures grown on high-density electrode arrays
Free-floating cerebral organoids in a hosted facility
Neuron–silicon cells packaged into sensing devices
Circuit connectivity
Specified as a design input, fabricated, then verified per batch
Largely emergent from culture conditions
Emergent within each organoid
Fixed by the device architecture
Structural control
Two-photon written 3D scaffolds define channels at sub-micron resolution
Planar substrate geometry
Self-organised organoid interior
Chip-level packaging
Engagement model
Wetware-as-a-service R&D programmes, with on-prem modules for partners
Benchtop unit plus cloud access
Remote subscription access to shared organoids
Sensor units sold for detection use cases
Primary application
CNS disorder modelling, drug discovery, and programmable biocomputing
Biological computing research
Low-energy biocomputing research
Chemical and olfactory detection
Autonomy roadmap
Prototype biohybrid control modules for UAV and defence integration
Not a stated focus
Not a stated focus
Detection payloads
Neural substrate
Biohybrid Devices
3D assembloids — organoid nodes fused through micropatterned axon tracts
Cortical Labs
Cortical neuron cultures grown on high-density electrode arrays
FinalSpark
Free-floating cerebral organoids in a hosted facility
Koniku
Neuron–silicon cells packaged into sensing devices
Circuit connectivity
Biohybrid Devices
Specified as a design input, fabricated, then verified per batch
Cortical Labs
Largely emergent from culture conditions
FinalSpark
Emergent within each organoid
Koniku
Fixed by the device architecture
Structural control
Biohybrid Devices
Two-photon written 3D scaffolds define channels at sub-micron resolution
Cortical Labs
Planar substrate geometry
FinalSpark
Self-organised organoid interior
Koniku
Chip-level packaging
Engagement model
Biohybrid Devices
Wetware-as-a-service R&D programmes, with on-prem modules for partners
Cortical Labs
Benchtop unit plus cloud access
FinalSpark
Remote subscription access to shared organoids
Koniku
Sensor units sold for detection use cases
Primary application
Biohybrid Devices
CNS disorder modelling, drug discovery, and programmable biocomputing
Cortical Labs
Biological computing research
FinalSpark
Low-energy biocomputing research
Koniku
Chemical and olfactory detection
Autonomy roadmap
Biohybrid Devices
Prototype biohybrid control modules for UAV and defence integration
Cortical Labs
Not a stated focus
FinalSpark
Not a stated focus
Koniku
Detection payloads
Comparison summarises each platform's publicly described positioning for orientation only. Other companies' capabilities evolve — please consult their current documentation. Company names are the trademarks of their respective owners.
Connectivity you can specify
The open gap across living-neuron platforms is control: cultures self-organise and every batch differs. Designed tracts make a circuit a spec you can restate, share and reproduce.
Built for pharma endpoints
Biocomputing-first platforms optimise for compute demos. We optimise for the endpoints CNS programmes are actually judged on — circuit-level phenotypes, dose response, isogenic controls.
A path off the bench
Research access and deployable hardware usually come from different vendors. The same assembloid architecture runs in our facility and inside a hardened control module.
04Who it is for
One platform, four kinds of programme.
Research access first — deployable hardware when your programme is ready for it.
Neuroscience research
Ask structural questions you could never control for in a dish.
Motif libraries: feed-forward, recurrent, and inhibitory-gated topologies
Plasticity and learning studies against a known wiring baseline
Longitudinal recordings across weeks, not single sessions
Biomedical engineering
A living test bench for interfaces, scaffolds and control laws.
Prototype neural interfaces against reproducible circuits
Co-design scaffolds with our two-photon fabrication team
Benchmark closed-loop controllers on real neural dynamics
Pharma & CNS discovery
Circuit-level endpoints for programmes that keep failing at cell level.
Patient-derived assembloids with isogenic controls
Chronic dosing schedules under continuous perfusion
Defence & autonomy
Adaptive control research, moving from bench to prototype module.
Prototype biohybrid control modules for UAV integration
Low-power adaptive control benchmarked against silicon baselines
Integration support for MoD programmes and prime contractors
Where this is heading
Revolutionising UAVs with living intelligence.
Biological circuits adapt, degrade gracefully and learn on a power budget silicon struggles to match. Our deployed control modules put an assembloid — with its perfusion, interface and safety envelope — into a hardened package that a defence integrator can actually mount. Today that work runs as funded R&D alongside prime contractors; the research platform is how programmes get there.
Indicative pricing shown for illustration. Deployed Control Module pricing reflects our current contract structure of $45,000 per month ($540K annual contract value per unit); research programmes are scoped per project. All figures on this page are sample data.
06Trajectory
From 3 modules in the field to 40 by Year 3.
Research programmes fund the platform; deployed control modules carry the contract value. Each module runs at $45,000 per month — $540K of annual contract value — reflecting what MoD programmes and primes pay for novel biocomputing IP plus the integration work around it.
Per-module ACV
$540K
Year 3 run-rate
$21.6M
Deployed control modules
Plan
3
Year 1
14
Year 2
40
Year 3
Live now — 3 modules under annual contract
Second production line, first prime-contractor programme
40 modules deployed · $21.6M run-rate at $540K ACV
Illustrative plan figures, shown as sample data.
07Voices from the bench
What partner teams say about designed connectivity.
Sample placeholder quotes, shown while our first cohort partners complete their publication and approval process.
“We had spent two years arguing about whether an effect was the drug or the culture. Being handed the connectivity map alongside the recordings ended that argument in a single cohort.”
IODr. Imani OseiPrincipal Investigator, Halden Institute for Neural Systems
“The first assembloid we specified came back wired the way we drew it. Our controller benchmarks finally have a substrate that does not drift between batches.”
RIRowan IyerHead of Neural Interfaces, Cambrian Autonomy Labs
“Circuit-level endpoints changed which compounds we advanced. Two candidates that looked identical on cell viability separated cleanly on burst structure.”
PRDr. Priya RaghunathanDirector of CNS Discovery, Meridian Neurotherapeutics
Placeholder testimonials from fictional organisations, shown for illustration.
08FAQ
Questions we get before the first call.
Anything else, and the fastest route is to put it in the waitlist form — a platform scientist answers directly.
It is a multi-node living circuit. Each node is a cerebral organoid; the nodes are joined by micropatterned neuron tracts that grow inside two-photon fabricated guidance channels. Because the channels are written before the tissue matures, the wiring diagram is something you specify up front rather than something you discover afterwards. Every build ships with the connectivity map it was manufactured against.
No. Research Access and Program Partner engagements run inside our facility: you design the protocol, we culture and fabricate, and you drive closed-loop sessions remotely through the research SDK. On-premise hardware only enters the picture for deployed control modules, and we commission those on site.
Those platforms all put living neurons to work, and they do it well. The difference is structural control: they largely rely on connectivity that emerges from culture or is fixed by a device, while we fabricate the topology as a specification and verify it per batch. We also lead with CNS disease modelling and drug-discovery endpoints rather than compute demonstrations, and we carry the same architecture into deployable control modules.
Our reference iPSC bank, isogenic control lines, or your own patient-derived lines transferred under a material transfer agreement. Line qualification runs as a short scoping phase before your first cohort so we can confirm organoid formation and baseline activity.
You own your protocols, your data and the discoveries that come from them. We retain the platform IP — fabrication methods, connectivity library and interface stack. Publication is encouraged; Program Partner engagements include a co-authorship pathway when our scientists contribute directly.
We work with iPSC-derived tissue under institutional oversight, within the size and complexity limits set by prevailing ethical guidance, and with no capacity for sentience claimed or sought. Every programme is reviewed against the governing framework in its jurisdiction, and defence work runs under the applicable export-control and programme rules.
Research Access cohorts are scheduled quarterly; the current queue is one cohort out. Custom connectivity designs for Program Partners add a design and fabrication cycle before first data. Deployed control modules are scoped individually, since integration timelines are set by the host platform.
Get started
Put your programme on the platform.
Tell us what you want to measure. A platform scientist reviews every request and comes back with a scoped cohort design — connectivity motif, cell source, endpoints and timeline — before anything is committed.
Scoping call with a platform scientist, not a sales rep
