All products

L4 / Implant · read

Reference code only

Reference code

Lattice

A record of a read interface.

Reference code for an implanted read-interface concept: a simulated motor decoder and a speech contract stub, kept outside the product line.

Reference code only. No clinical product or service is offered or planned.

Discuss the record
01The record

Record the interface; claim no decoding.

Lattice is the July 2026 concept of an implanted read interface for deliberate communication and creation: a conforming cortical leaf and a crescent processing spine, drawn so that the person’s intended message, timing and ability to cancel would lead the design. Its registry status is reference. The tree, the design record and this page stay so the record is whole; no device, clinical product or service is offered or planned from it.

Registry status
Reference · outside the product line
Concept
Implanted read interface · cortical leaf and spine
Implemented reference
Simulated motor decoder · speech contract stub
Data boundary
Raw neural processing inside the implant; features only leave
02The physical design

A recording leaf, open to the surface.

A curved cortical leaf spreads into rounded apertures and independently conforming petals. A crescent spine gathers acquisition and processing at its edge, paired with a small external hairline saddle.

Lattice anatomy showing the open recording leaf, separated crescent processing zones and external saddle cap, component zones, tapered carrier and removable contact skirt.
AnatomyResearch concept — hardware unvalidated

44 × 28 × approx. 7 mm

The external coupler is a fit target. The enlarged cortical leaf and crescent spine are schematic, with no anatomical dimensions claimed. Coupling and chronic packaging require clinical-partner development.

  1. Give the recording surface room.

    An open cellular leaf and several petals represent the cortical recording branch. Sparse visible sites explain their role; they do not stand for an established channel count, anatomical fit or chronic signal quality.

  2. A spine with a clear task.

    The edge crescent reserves acquisition, implant-local feature processing, and derived-data link and power zones. Lattice is a read-only interface; all raw neural processing remains inside the implant boundary.

  3. Shape the part you handle.

    A mineral external saddle meets a replaceable graphite contact skirt and tapered flexible carrier. Its keyed desktop cradle explores easy reach and alignment, including assisted use.

03Why it is registered

Why the record is kept.

A registry row, a devices tree and a catalog entry are one identity (ADR-0037), so the concept is kept as reference code rather than deleted. Lattice is not part of the company story: Cr4dle is a non-commercial research programme now and plans to offer in a later phase, behind dated gates, collections of consented, labelled episodes recorded on the capture kit on bought sensors, the kit licence and dataset licences (nothing is offered for sale today), and no implanted interface, communication product or clinical service is planned. Reliable activation, correction and silence were the concept’s clinical research targets; they stay in the record as questions a specialist partner would have to answer under design controls, not as claims.

Our approach to evidence
04Work in progress

Where we are.
What comes next.

Implemented reference: a simulated motor decoder, local feature-export boundaries and a speech contract stub, retained as is. There is no working speech decoder and no implanted hardware; durable communication and an expressive voice would need partner-led engineering, design controls and clinical evaluation, none of which is pursued.

A clear boundary

An intended communication channel is not unrestricted access to a person’s thoughts. Reference code establishes no capability; all Lattice hardware and clinical outcomes remain unvalidated.

Recorded direction, not pursued (reference code only)

Designed forExpression on your terms.
Communicate in a chosen voice, correct a message and create in a personal workspace, with explicit activation and a dependable way to remain silent.
Next proofA complete communication day.
Clinical partners must test useful communication across real sessions, including activation errors, correction, fatigue, setup and calibration drift.
Research horizonDeliberate composition through inner speech.
Research deliberately gated inner speech and personally calibrated retrieval. Require evidence that the person can reliably initiate, revise and stop every proposed output.
Continue exploring

See the wider picture.

The kit families, the research tiers and the reference designs each have their own evidence path.

Compare the families