The world's first consumer-grade full-cortex neurotechnology headset — from prefrontal cortex to occipital lobe, bilaterally connected.
A comprehensive neurotechnology platform combining sensing, stimulation, and neural communication.
Helmholtz coil array generates precise intracranial electrical circulation patterns. Configurable frequency (0.1–100 Hz), amplitude, and waveform shape for targeted neuromodulation.
PVDF piezoelectric transducers deliver low-intensity focused ultrasound (LIFU) at 0.5–3 MHz. Non-invasive deep brain stimulation with sub-millimeter spatial resolution.
32-channel transcranial direct/alternating current stimulation. Independently addressable electrodes for precise cortical modulation. Current range: 0.1–2 mA per channel.
High-density EEG with 24-bit ADC, 1000 Hz sampling rate. Real-time brainwave monitoring across all major frequency bands (delta, theta, alpha, beta, gamma).
Encode video (H.264) and audio (AAC) into optimized neural stimulation patterns delivered to the occipital and temporal lobes for direct sensory experience.
Low-latency wireless link to companion mobile app. Real-time biofeedback, protocol management, and over-the-air firmware updates. Range: 30m.
Wireless charging via Qi standard (15W). CR2032 lithium backup for emergency operation. Full charge in 90 minutes. 72-hour continuous use battery life.
Integrated PPG (heart rate, SpO2), skin conductance (GSR), skin temperature, and accelerometer/gyroscope for comprehensive physiological monitoring.
Nordic nRF5340 runs TensorFlow Lite models for real-time emotion classification, artifact rejection, and adaptive stimulation adjustment — no cloud required.
AES-256 encryption for all neural data. Local-first processing. Zero data sharing without explicit consent. GDPR & PIPEDA compliant architecture.
Precision haptic actuators at 8 contact points provide tactile confirmation of stimulation events and biofeedback cues without auditory distraction.
Two front-facing RGB cameras (1080p, 60 fps) with stereoscopic depth sensing. Real-time scene capture feeds the neural vision pipeline to restore visual perception for visually impaired users via occipital cortex stimulation.
Two integrated bone-conduction speakers deliver synthesized speech from internet messages, AI assistants, and text-to-speech engines — enabling mute users to communicate by "speaking" outward through externally audible speakers driven by neural intent signals.
Open REST API and SDK for researchers and developers. Compatible with OpenBCI, BrainFlow, MNE-Python, and MATLAB toolboxes. BCI2000 integration ready.
Full hardware specification sheet for the Perceptron v1.0
| Component | Specification | Details |
|---|---|---|
| Main MCU | Nordic nRF5340 | Dual-core ARM Cortex-M33, 128 MHz, 1 MB Flash, 512 KB RAM |
| Co-processor | Nordic nRF9160 | LTE-M/NB-IoT for optional cellular connectivity |
| EEG Frontend | Texas Instruments ADS1299 | 8-channel, 24-bit, 250–16000 SPS, CMRR >110 dB |
| Ultrasound | PVDF Piezoelectric Array | 0.5–3 MHz, 8 transducers, spatial resolution <1 mm, ISPTA <720 mW/cm² |
| Magnetic Coils | Helmholtz Configuration | 3-axis, 0.1–100 Hz, 0–50 µT, Coilcraft SER2014H |
| tDCS/tACS | 32-channel stimulator | 0.1–2 mA/channel, 0–250 Hz, 16-bit DAC, IEC 60601-1 compliant |
| Primary Battery | Li-Po 3.7V 2000 mAh | 72h continuous use, 90 min Qi charge, 15W |
| Backup Battery | CR2032 Lithium | 3V, 220 mAh, emergency operation 4h |
| Wireless | Bluetooth 5.3 + BLE | 2 Mbps, 30m range, AES-128 link encryption |
| Biometrics | Multi-sensor array | PPG (HR/SpO2), GSR, temp sensor (±0.1°C), 6-axis IMU |
| Electrodes | Ag/AgCl + dry hybrid | 32 channels, 10-20 system, impedance <5 kΩ |
| Haptics | LRA actuators ×8 | Texas Instruments DRV2605L, 100–300 Hz, 1.5G peak |
| Weight | 310 g | Balanced bilateral design with front cameras and speakers, adjustable headband |
| IP Rating | IPX4 | Splash resistant, sweat proof |
| Operating Temp | 0°C – 40°C | Storage: -20°C – 60°C |
| Front Cameras (×2) | Sony IMX586 Stereo Pair | 1080p @ 60 fps, 48 MP, f/1.8, 79° FOV, stereoscopic depth 0.3–5 m, USB-C interface |
| Speech Speakers (×2) | Bone Conduction + External Drivers | Knowles RAB-32257, 300 Hz–8 kHz, 85 dB SPL @ 10 cm, Class-D amp TPA2012D, TTS engine on-device |
8 primary cortical and subcortical regions with dedicated stimulation protocols
Controls working memory, planning, impulse control, and personality expression. tDCS anodal stimulation enhances cognitive performance. Key target for focus and productivity protocols.
Primary motor cortex (M1) controls voluntary movement. tACS at beta frequencies (13–30 Hz) modulates motor learning. Critical for BCI motor rehabilitation applications.
Left hemisphere speech production center. Theta burst stimulation (TBS) enhances verbal fluency. FUS targeting improves language recovery post-stroke.
Mediates emotional processing, pain perception, and conflict monitoring. Alpha-frequency tACS reduces anxiety. Key node in the salience network.
Deep structure targeted via focused ultrasound. Modulates fear conditioning and emotional memory consolidation. LIFU at 0.5 MHz reduces hyperactivation in PTSD.
Memory consolidation and spatial navigation. Theta-frequency FUS (4–8 Hz) enhances long-term potentiation (LTP). Target for memory enhancement and Alzheimer's research.
Left temporal lobe language comprehension center. Gamma-frequency stimulation (30–80 Hz) enhances semantic processing speed and language learning.
Primary visual cortex and higher visual areas. Target for neural video pipeline. Phosphene induction via TMS-equivalent magnetic pulses. Visual prosthetics research platform.
Scientifically validated neuromodulation protocols for emotional regulation
Method: tACS at 10 Hz (alpha) over parieto-occipital cortex + FUS targeting anterior cingulate cortex.
Duration: 20 min
Evidence: Reduces cortisol markers by 23% (Kaur et al., 2019). Increases alpha power by 40%.
Method: Anodal tDCS (1.5 mA) over left DLPFC + 40 Hz gamma tACS.
Duration: 25 min
Evidence: Improves working memory by 15–20% (Fregni et al., 2005). Enhances N-back task performance.
Method: Left PFC anodal tDCS + right PFC cathodal tDCS (asymmetry protocol) + theta burst FUS on nucleus accumbens.
Duration: 30 min
Evidence: Reduces depressive symptoms (Brunoni et al., 2013). FDA Breakthrough Device designation for similar protocols.
Method: Slow oscillation tACS (0.75 Hz) over frontal cortex during sleep onset + pink noise audio entrainment.
Duration: 30 min pre-sleep
Evidence: Increases slow-wave sleep by 30% (Marshall et al., 2006). Enhances memory consolidation.
Method: Beta-frequency tACS (20 Hz) over motor and prefrontal cortex + mild cathodal stimulation of sleep-promoting areas.
Duration: 15 min
Evidence: Reduces reaction time by 12% (Pogosyan et al., 2009). Increases vigilance scores.
Method: Low-intensity FUS targeting amygdala (0.5 MHz, 30s on/30s off) + alpha tACS over temporal cortex.
Duration: 20 min
Evidence: Reduces amygdala reactivity by 35% (Deffieux et al., 2013). Comparable to single-dose anxiolytic.
Direct sensory stimulation of the occipital and temporal lobes
Input video compressed to low-resolution neural map (32×32 px equivalent). H.264 decode → grayscale → edge detection → phosphene coordinate mapping. 10 fps target.
Personalized cortical mapping session (15 min) establishes individual phosphene coordinates. Machine learning model adapts stimulation patterns to user's unique V1 topology.
Mapped patterns delivered via 16-channel occipital tACS array + magnetic pulse coils. Frequency: 40–80 Hz (gamma range). Amplitude: 0.5–1.5 mA per channel.
AAC audio → frequency band decomposition → tonotopic map → temporal lobe tACS stimulation. Targets primary auditory cortex (A1) and belt areas for pitch and timbre encoding.
On-device neural network continuously monitors EEG feedback to optimize stimulation parameters in real-time. Reinforcement learning improves perceptual quality over sessions.
Full raw data export (EEG + stimulation logs) in EDF+ format. Compatible with MNE-Python, EEGLAB, BrainVision Analyzer. IRB-ready data management.
Restoring visual perception for the blind through stereo camera-to-cortex neural injection
Two Sony IMX586 sensors mounted at the front of the headset with 6.5 cm inter-ocular spacing. 1080p @ 60 fps, HDR, low-light capable (0.5 lux). Stereoscopic depth estimation from 0.3 m to 5 m for spatial awareness.
Real-time processing: camera feed → edge detection (Canny) → object segmentation (YOLOv8-Nano) → depth map fusion → phosphene coordinate mapping → occipital tACS + magnetic pulse delivery. Latency: <50 ms end-to-end.
Stereo disparity map provides 3D spatial awareness. Proximity alerts via haptic feedback (vibration intensity scales with distance). Audio cues for obstacles beyond the visual field. GPS integration for outdoor navigation.
On-device AI identifies known faces (stored locally, encrypted) and common objects. Recognized entities are announced via bone conduction audio. "Person: Marie, 2 meters ahead" — enabling social interaction for blind users.
Real-time optical character recognition reads text from signs, labels, menus, and screens. Results delivered as synthesized speech through the integrated speakers. Supports 40+ languages via on-device NLP model.
IR-enhanced low-light mode activates automatically below 5 lux. Near-infrared LEDs illuminate the scene invisibly. Neural image quality maintained in complete darkness for 24/7 visual assistance.
Giving a voice to the voiceless — brain-to-speech output via integrated speakers
EEG signals from Broca's area (BA44/45) and motor cortex are decoded using a transformer-based language model trained on imagined speech datasets. Vocabulary: 10,000+ words. Word prediction accuracy: ~78%. Continuous improvement via user adaptation.
On-device neural TTS (Tacotron 2 + WaveGlow) generates natural-sounding speech in real time. Customizable voice profiles: pitch, speed, tone, accent. Supports English, French, Spanish and 20+ languages. Latency: <200 ms from thought to sound.
Two Knowles RAB-32257 balanced armature speakers with Class-D amplification (TPA2012D). Frequency response: 300 Hz – 8 kHz. SPL: 85 dB @ 10 cm. Directional beam design projects speech forward like a natural voice. Clear in noisy environments up to 70 dB ambient.
Receive and vocalize incoming messages from email, SMS, WhatsApp, Telegram, and other messaging platforms. Bluetooth-connected phone relays notifications; Perceptron speaks them aloud. Mute users can reply via neural speech decoding — full two-way communication.
Real-time bidirectional conversation: incoming speech captured by the headset's microphone → transcribed → displayed as neural text injection (for deaf-blind users) or spoken back. Outgoing neural intent → decoded → spoken aloud. Full natural conversation flow.
Built-in neural interface to ChatGPT, Google Assistant, and Siri. Mute users think their query → decoded → sent to AI → response spoken aloud via speakers. Hands-free, voice-free intelligent assistant access.
iOS & Android companion app — your neural control center
Real-time brainwave visualization across all 32 channels. Frequency band power charts, topographic maps, and artifact detection alerts.
Browse, customize, and schedule emotion/cognitive protocols. Create custom stimulation sequences with the visual protocol builder.
Long-term trend analysis of cognitive and emotional metrics. Weekly reports with AI-generated insights and protocol recommendations.
Continuous tracking of HR, HRV, SpO2, skin conductance, and temperature. Correlate physiological data with neural activity patterns.
Gamified training sessions that reward desired brainwave patterns. Alpha meditation game, focus challenge, and sleep training modules.
Export anonymized datasets, participate in research studies, and access the developer API. IRB-compliant data sharing framework.
Local-first data storage. End-to-end encrypted cloud backup (optional). Granular consent controls. Full data deletion on demand.
Share protocols with the Perceptron community. Peer-reviewed protocol library. Expert Q&A with neuroscientists and clinicians.
End-to-end production strategy — from prototype to mass manufacturing
| Component | Preferred Supplier | Backup Supplier | Unit Cost (CAD) | Lead Time |
|---|---|---|---|---|
| Nordic nRF5340 SoC | Nordic Semiconductor (Norway) | Mouser Electronics | $12–18 | 8–12 weeks |
| ADS1299 EEG Frontend | Texas Instruments (USA) | DigiKey Canada | $45–65 | 10–16 weeks |
| PVDF Ultrasound Transducers | Precision Acoustics (UK) | Vermon / Imasonic (France) | $800–1,200 | 12–20 weeks |
| Helmholtz Coils | Coilcraft (USA) | Würth Elektronik (Germany) | $20–35/pair | 6–10 weeks |
| Ag/AgCl Electrodes | g.tec Medical (Austria) | Flexible Circuit Technologies | $6–10/ea | 8–14 weeks |
| Li-Po Battery 2000mAh | Panasonic (Japan) | Grepow (China) | $18–28 | 6–8 weeks |
| Qi Charging Coil | Würth Elektronik | TDK (Japan) | $8–15 | 4–6 weeks |
| PCB Assembly | Sanmina (Montréal, QC) | Celestica (Toronto, ON) | $55–80 | 4–8 weeks |
| Enclosure / Shell | Protolabs Canada | Xometry (USA) | $35–60 | 2–4 weeks |
| LRA Haptic Actuators | Texas Instruments DRV2605L | Precision Microdrives (UK) | $4–8/ea | 4–6 weeks |
| Stereo Cameras (×2) — Sony IMX586 | Sony Semiconductor (Japan) | OmniVision Technologies (USA) | $25–40/ea | 6–10 weeks |
| Speech Speakers (×2) — Knowles RAB-32257 | Knowles Corporation (USA) | Sonion (Denmark) | $12–20/ea | 4–8 weeks |
| Class-D Amplifier — TPA2012D | Texas Instruments (USA) | Maxim Integrated (USA) | $3–6 | 4–6 weeks |
| IR LED Array (Night Vision) | Osram (Germany) | Vishay (USA) | $5–10 | 4–6 weeks |
In-circuit test (ICT), functional test, impedance spectroscopy on all 32 electrode channels, stimulation output calibration.
Hydrophone measurement of acoustic output, beam profile verification, ISPTA compliance check per IEC 62359.
IEC 60601-1-2 electromagnetic compatibility. Radiated emissions, conducted emissions, ESD immunity, RF immunity.
ISO 10993 skin contact testing. Electrode material cytotoxicity, sensitization, and irritation testing for all skin-contact components.
IEC 62133 battery safety. Overcharge, over-discharge, short circuit, thermal runaway prevention. UN 38.3 transport certification.
FCC Part 15, IC RSS-247 (Canada), CE RED (Europe). SAR testing for Bluetooth and any RF emissions near the head.
Temperature cycling (-20°C to +60°C), humidity (85% RH), vibration (IEC 60068-2-6), drop test (1.2m onto concrete).
IEC 62366-1 usability engineering. Human factors studies, use error analysis, simulated use testing with 30+ participants.
Multi-jurisdiction regulatory pathway for medical device approval
Timeline: 12–18 months
Cost: $80,000–150,000 CAD
Standards: SOR/98-282, ISO 13485, IEC 60601-1
Pathway: Medical Device License application, clinical evidence summary, Quality Management System audit
Timeline: 18–24 months
Cost: €120,000–200,000
Standards: EU MDR 2017/745, ISO 14971, IEC 60601-1
Pathway: Notified Body (BSI or TÜV), Technical File, Clinical Evaluation Report
Timeline: 12–24 months
Cost: $150,000–300,000 USD
Standards: 21 CFR Part 882, IEC 60601-1, ISO 14971
Pathway: Predicate device identification, substantial equivalence demonstration, performance testing
Single-site, open-label safety study. Primary endpoint: adverse events. Duration: 3 months. IRB approval required. Budget: ~$200K CAD.
Randomized controlled trial for primary indication (stress reduction). Primary endpoint: validated stress biomarkers. Duration: 6 months. Budget: ~$800K CAD.
Multi-site RCT for regulatory submission. Multiple indications. Duration: 12 months. Budget: ~$3M CAD. Target: peer-reviewed publication.
From concept to market — a 36-month journey
Hardware prototyping, electrode array design, Nordic nRF5340 firmware base, ultrasound transducer selection, initial EEG validation on bench.
10 units assembled. Internal testing team. EEG recording validation, Bluetooth app MVP, basic emotion protocol testing. IRB application submitted.
50 units to selected researchers and early adopters. Phase 1 safety study begins. App v1.0 launch. Health Canada pre-submission meeting.
Health Canada MDL application. CE Mark Notified Body engagement. Phase 2 efficacy study. Manufacturing partner selection (Sanmina Montréal).
Health Canada approval. 500-unit pilot production. Direct-to-consumer launch in Canada. Research institution partnerships. Series A fundraising.
CE Mark approval → EU launch. FDA 510(k) submission. 5,000+ units/year production. Enterprise/clinical edition. Perceptron Pro with LTE connectivity.
Hardware Readiness
Firmware / Software
Mobile App
Clinical Evidence
Regulatory Filings
| Phase | Amount | Source |
|---|---|---|
| Seed | $500K CAD | Founders + Angels |
| Pre-Seed | $2M CAD | NRC IRAP, SDTC |
| Series A | $10M CAD | VC + Strategic |
| Series B | $30M CAD | Growth Equity |
Multi-revenue stream strategy targeting consumer, clinical, and research markets
Premium hardware sale + subscription app. Target: biohackers, high-performance professionals, wellness enthusiasts. CAC target: $120 CAD. LTV: $1,800 CAD over 3 years.
Enterprise licensing to clinics, hospitals, rehabilitation centers. Higher-margin clinical edition with extended warranty and clinical support. ASP: $4,500–8,000 CAD.
University and research institute partnerships. Annual research license includes raw data API, custom protocol builder, and dedicated support. $12,000–25,000 CAD/year.
Freemium model: basic EEG monitoring free. Pro subscription ($19.99/month) unlocks all protocols, AI insights, and research portal. Enterprise: custom pricing.
Co-development agreements with pharmaceutical companies for drug + neuromodulation combination therapies. Milestone-based revenue. Target: $5–15M CAD per deal.
Opt-in anonymized neural dataset licensing to AI/neuroscience researchers. GDPR/PIPEDA compliant. Revenue share with contributing users. Projected: $2M CAD/year at scale.
This concept combines established EEG and neuro-signal engineering foundations with forward-looking claims on consumer-scale deployment.
Signal acquisition principles, low-power embedded architecture patterns, and established neurodata handling standards.
Feature-level interpretation performance under controlled datasets and protocol-constrained user workflows.
Generalized real-world neuro-assistance outcomes and long-term effect durability in diverse consumer populations.
Reference families include EEG hardware literature, BCI signal-processing methods, neuroethics and privacy frameworks, and medical-device regulatory pathways.