CogSoma · by cagelab

Every rig
within reach.

A toolkit that makes the neuroscience lab automated and digital, so people and AI can sense, operate and manage experiments directly.

What CogSoma isGet started
What CogSoma is

Run. Record. Reach.

Experiments run on their own in the cage. Everything they produce is recorded around the animal. People and AI reach all of it through the same interfaces.

01 · RUNExperiments run themselvesRigs in and around the cage that run experiments on their own.
CageLabLow-cost in-cage touchscreen training kiosk with pump, feeder, camera and speaker modules.MXBImxbiflow: multi-animal, multi-stage touchscreen experiments with RFID, rewarder and detector drivers.cagelab-2Modular real-time behavioral control core on Raspberry Pi or x86 controllers.
PLANNEDClassic instrumentsStandard lab equipment and programs, such as Open Ephys.
cogmoteGORuns on every rig: one remote interface for status, video, services, power and commands to the task.THE INTERFACE TO EVERY RIGpymoteGO · matmoteGOPython and MATLAB task code talks to the rig.
02 · RECORDThe lab, as dataOne record for everything the lab does.
CogMatrix

An animal-centred database: each animal’s lifecycle and care, its sessions, the devices and protocols behind them, and the data files and datasets they produce.

03 · REACHOpen to people and AIThe same doors for every user.
cogmoteHub that knows every rig by its device code: enrollment, config, heartbeat, notifications and access grants. Web app and CLI.CogIDOne identity for people, AI and task scripts.

People and AI, as equal users.

PeopleResearchers and operators use the web apps, the cogmote CLI and these docs.AIWorks from its own computer through the cogmote CLI, its skill and the Python SDK.

same CogID sign-insame cogmote hubsame CogMatrix record

DIRECTION

AI is a first-class user. We keep adding interfaces so AI can do, and do better, what only people could do in the lab.

How we got here

From one automated cage to a lab AI can use.

  1. 01Automated cagesIn-cage rigs, CageLab and MXBI. cogmoteGO automated the experiments on them.
  2. 02Managing the rigsA pipeline to manage the instruments became the cogmote hub.
  3. 03One interface to every rigcogmoteGO grew into a general remote interface: run experiments and collect data on all kinds of rigs.
  4. 04A digital labCogMatrix grew from serving automation into an animal-centred record of the whole lab.
  5. 05AI arrivesAI became a natural user of the system, next to people.
  6. NEXT · PLANNEDClassic instrumentsBring standard equipment and programs, such as Open Ephys, within reach.

Start where you work.

ResearchersRecord animals, run sessions, find data.
OperatorsDeploy the services and keep rigs running.
Developers & agentsBuild on the APIs, or hand them to an agent.

References

GRAPHQLCogMatrix APIMagidocDATABASEService databasestbls + Liam ERDPYTHONCogMatrix SDKpdocRESTcogmote hub APIScalarRESTcogmoteGO device APIScalar

Every rig within reach.

For people and AI alike.