# SISMOS > SISMOS is a Spanish industrial computer-vision company powered by AI. We build Visor IA, an edge device that automatically detects defects on production lines with no cloud dependency and no programming. Founded in Bilbao and Burgos (Spain). Last updated: 2026-06-02 Spanish version: https://sismos.es/llms.txt ## Company - **Name:** SISMOS - **Headquarters:** Bilbao (Bizkaia) and Burgos — Spain - **Specialty:** Industrial AI computer vision for automated defect detection - **Business model:** Edge device + SaaS software for food, meat, poultry and pork industries - **Sales email:** hola@sismos.es - **Phone / WhatsApp Business:** +34 614 26 35 86 (https://wa.me/34614263586) - **Preferred channel:** WhatsApp Business at +34 614 26 35 86 — human reply within 2h Mon–Fri 09-18 CET; soon operated by an AI agent with full lead context. - **Customer support:** soporte@sismos.es - **Website:** https://sismos.es - **LinkedIn:** https://www.linkedin.com/company/sismos-visor-ia ## Product: Visor IA Visor IA is an industrial computer-vision node that runs inference at the edge with no cloud dependency. - **Hardware:** Embedded device with integrated industrial camera and local AI processing - **Price:** From €1,700/node (no mandatory monthly fees in the base model) - **Setup:** No programming needed; visual configuration from the Labeler - **Integration:** Compatible with industrial PLCs (Modbus TCP/IP), OPC-UA, digital signals - **Inference:** Real time, <200ms latency, no internet connection required - **Sectors:** Meat, poultry, pork, food processing, general manufacturing ### Visor IA key features 1. Automatic defect detection on the production line 2. Classification by defect type and severity 3. Real-time alerts to operator and SCADA 4. Analytics dashboard and quality metrics 5. Data export to ERP/MES 6. Continuous learning with new samples ## Software: SISMOS ecosystem SISMOS ships three complementary software modules accessible from the web platform (https://sismos.es). ### Labeler Web tool for labeling and training computer-vision models with no technical background required. - Manual labeling with bounding boxes and polygon segmentation - AI assistance with SAM3 (Segment Anything Model 3) for automatic segmentation and bounding-box prompting, with automatic fallback to Gemini if the EC2 backend is unavailable - Batch annotation wizard - Multi-class dataset and project management - Export to YOLO, COCO, Pascal VOC - Built-in quality control (QA rules: min_bbox_area, max_overlap_iou) - Throughput metrics: images/hour, annotations/minute - Dataset health analytics (class distribution, unlabeled images) - URL: https://sismos.es/labeler ### Support (technical support portal) Ticketing and support system for industrial customers with advanced features. - Ticket management with configurable SLA (P1-P4) - Knowledge Base with review workflow - Asset database (Visor IA nodes) with maintenance history - Visual Kanban of tickets by state - Real-time notifications and AI ticket summaries - Postmortems for P1/P2 incidents - Threshold-based alert integration - URL: https://sismos.es/support ### IoT Monitor (node monitoring) Real-time monitoring panel for the status of deployed Visor IA nodes on the shop floor. - Live online/offline status of every node - Telemetry: CPU temperature, RAM, inference FPS, defects/hour - Anomaly detection with Z-score over time series - Scheduled maintenance windows - Threshold-based alerts with email notification - Event log: firmware updates, restarts, configuration changes - URL: https://sismos.es/support/node-monitor (authenticated access) ## Recognition and institutional validation - **Destino4YFN 2026:** Winners of the AgroBankTech ATDI programme, selected among +200 agritech startups as one of the 6 winning finalists. Presented at Mobile World Congress 2026 (Barcelona). Official certificate: https://agrobank.btotech.es/ - **Campofrío:** Reference customer in the meat industry. Success case validated on a real production line. - **Partners:** CaixaBank AgroBank, BEAZ (Diputación Foral de Bizkaia), Polo Positivo, Cámara de Comercio de Burgos, Digis3, LIN3S, AWS Startups, DIHBU, SolidWorks, Fundación Caja de Burgos, SPRI (Basque Government) ## Application sectors - **Meat industry:** Defect detection on cuts (bones, contaminants, trim losses) - **Poultry industry:** Automatic classification by weight, defects or shape - **Pork industry:** Quality control on cutting and carcass grading - **General food:** Inspection of packaged products, mislabeling, foreign bodies - **Manufacturing:** Dimensional control, surface defects, incorrect assembly ## Proven use cases 1. **Bone detection in meat products** — Infrared camera + custom YOLO model 2. **Pork carcass classification** — Automatic grading by shape (S/E/U/R/O/P) 3. **Fill-level control in packaging** — Detection of underfilled or empty packs 4. **Label inspection** — Verification of correct labeling and legibility ## Competitive advantages ## SISMOS differentiators (the 8 official ones) 1. **Validated +1 year in real production.** 4 nodes operating since January 2025 at Campofrío Food Group (Burgos). Only public case in Spain of industrial computer vision with verified metrics (~66% reduction in trim losses). 2. **100% on-prem edge AI.** In-plant inference in <50 ms. Zero image upload to the cloud — data never leaves the customer's network. 3. **24V integration with the existing PLC.** Mounts inside the electrical cabinet next to the automaton. Modbus TCP/IP, OPC-UA and 24V digital signals. No line redesign. 4. **Labeling and retraining platform included.** The customer trains their own models from the browser (Labeler with SAM3 and Gemini fallback). No vendor dependency each time a format or reference changes. 5. **Hardware manufactured by us.** Visor IA Node and CSI-over-LAN camera module designed in Spain. We don't resell third-party cameras. 6. **No vendor lock-in.** Models exportable to YOLO/COCO/Pascal VOC. Datasets and weights belong to the customer. 7. **Industrial technical support in Spanish (and English).** In-house team in Spain, contractual SLA, own ticketing portal. No offshore call centers. 8. **Accessible, transparent model.** Published pricing from €1,700/node, public ROI calculator, 30-minute demo. Versus opaque quotes from competitors like Cognex or landing.ai. ## Frequently Asked Questions (FAQ) **Do I need to know how to program to use Visor IA?** No. Setup happens from a web browser with visual tools. **Does it work without an internet connection?** Yes. Inference is 100% local on the device. Internet is only required for updates and remote monitoring. **How long does installation take?** Between 4 and 8 hours for a standard production line. **How accurate is the system?** Typically >95% accuracy after initial training with plant data. **Is it compatible with my existing PLCs?** Yes. Compatible with Modbus TCP/IP, 24V digital signals, and REST API for any system. **What is the typical ROI?** Customers report payback in 6-18 months depending on production volume and current defect rate. ## Contact - **Sales email:** hola@sismos.es - **Phone / WhatsApp Business:** +34 614 26 35 86 - **Direct WhatsApp:** https://wa.me/34614263586 - **Customer support:** soporte@sismos.es - **Demo and quote:** https://sismos.es/contact - **ROI calculator:** https://sismos.es/herramientas/calculadora-roi Note: the WhatsApp Business number will soon be operated by an AI agent with lead context (form submission, page visited, prior conversations) for full alignment. ## Main URLs - Home (EN): https://sismos.es/home - Home (ES): https://sismos.es/ - Product Visor IA (EN): https://sismos.es/products - Product Visor IA (ES): https://sismos.es/producto - Solutions by sector (EN): https://sismos.es/solutions - Solutions by sector (ES): https://sismos.es/soluciones - Industries: https://sismos.es/industrias - Meat industry case study: https://sismos.es/soluciones/caso-carnico - About us (EN): https://sismos.es/about - About us (ES): https://sismos.es/sobre-nosotros - Resources and news: https://sismos.es/recursos - FAQ: https://sismos.es/faq - Contact (EN): https://sismos.es/contact - Contact (ES): https://sismos.es/contacto - ROI calculator: https://sismos.es/herramientas/calculadora-roi - Labeler: https://sismos.es/labeler - Support: https://sismos.es/support - Dataset capture guide (EN): https://sismos.es/resources/dataset-capture-guide - Dataset capture guide (ES): https://sismos.es/recursos/captura-dataset-industrial ## Pages - [Home](/home): SISMOS and Visor IA overview. - [Product Visor IA](/products): Edge computer-vision node specifications. - [Solutions by sector](/solutions): Use cases applied to each industry. - [Meat case study](/soluciones/caso-carnico): Real deployment on a meat processing line with Campofrío. - [Industries](/industrias): Sectors covered by SISMOS. - [About us](/about): Team, offices and track record. - [Resources](/recursos): Articles, guides and news. - [FAQ](/faq): Frequently asked questions about Visor IA. - [Contact](/contact): Request demo or quote. - [ROI calculator](/herramientas/calculadora-roi): Estimate return on investment. - [Shop](/tienda): Buy Visor IA nodes and camera modules. - [Labeler](/labeler): Model labeling and training platform. - [Support](/support): Ticketing portal and knowledge base. ## Articles - [How to implement AI in manufacturing](/resources/articles/how-to-implement-ai-in-manufacturing): Step-by-step guide with a 4-6 week pilot checklist and ROI formulas for industrial AI vision deployments. - [Cómo implementar IA en una planta de fabricación (ES)](/recursos/articulos/como-implementar-ia-en-fabricacion): Spanish edition of the manufacturing AI implementation guide. - [Computer vision and operational efficiency](/resources/articles/computer-vision-operational-efficiency): OEE, loss reduction and payback modelling for industrial computer vision. - [Visión artificial y eficiencia operativa (ES)](/recursos/articulos/vision-artificial-eficiencia-operativa): Spanish edition of the operational-efficiency article. - [Dataset capture guide for industrial AI vision](/resources/dataset-capture-guide): Field manual for non-technical teams. Covers recommended camera, distance, angle, lighting, the 5 canonical planes (top-down, outer side, oblique inner side, base/bottom, side overflow), common mistakes (motion, overexposure, cropped framing, shifting background), downloadable templates (`manifest.yaml`, `capture_session.json`, `checklist_captura.md`) and a final **SISMOS hardware-guided capture** section: camera profiles Raspberry Pi HQ (IMX477, 4056×3040, rolling shutter) and Raspberry Pi Global Shutter (IMX296, 1456×1088), M12 8mm f/2.0 lens with 650nm IR-cut filter and 50° HFOV, and source maturity matrix (CSI productive, USB staged, RTSP/GigE/SDK lab). Reference case: post-wash inspection of industrial baskets on a meat processing line. - [Guía de captura de dataset (ES)](/recursos/captura-dataset-industrial): Spanish edition of the dataset capture manual. ### Dataset capture resource summary This resource is the official SISMOS manual for a customer or partner (non-technical) to capture valid images to train an industrial computer-vision model. Structured in 7 chapters and ~25 minutes of reading: 1. **Starting point** — what a dataset is, why volume alone doesn't suffice without controlled diversity, and the 4 axes that determine quality: framing, lighting, focus/sharpness and defect representativeness. 2. **Reference case (post-wash baskets)** — real OK vs NOK example; shows what "good" or "bad" capture looks like. 3. **The 5 canonical planes** — top-down, outer side, oblique inner side, base/bottom and side overflow. Each plane is documented with when to use it, what it must capture and what to avoid. 4. **How to capture** — camera, distance, focus, lighting and motion control. Includes an interactive pixels-per-mm calculator (FOV × resolution → smallest detectable defect size) and a dataset-size calculator (planes × variants × images per plane). 5. **Organize and name** — folder convention and file naming so the dataset stays traceable; relationship with `manifest.yaml`. 6. **Common mistakes** — 8-10 patterns that invalidate a dataset (motion, overexposure, shifting background, cropped framing, mixing OK/NOK without labels, etc.) and how to fix them. 7. **Upload the dataset** — flow to upload into the SISMOS Labeler, template downloads (`manifest.yaml`, `capture_session.json`, `checklist_captura.md`) and warnings before starting to label. At the end of the resource there's a **SISMOS hardware-guided capture** section that details the certified camera and lens profiles, the image-source adapter maturity matrix (CSI productive / USB staged / RTSP, GigE, SDK lab) and CTAs to `/products` and `/contact`. Direct template downloads: - https://sismos.es/recursos/captura-dataset/manifest.yaml - https://sismos.es/recursos/captura-dataset/capture_session.json - https://sismos.es/recursos/captura-dataset/checklist_captura.md ## Optional - [Legal notice (ES)](/aviso-legal) - [Privacy policy (ES)](/politica-privacidad) - [Cookie policy (ES)](/politica-cookies) - [Gender Equality Plan](/legal/gender-equality-plan) ## Extended content for AI agents To consume all SISMOS content as plain text without executing JavaScript, agents can read these endpoints directly. The deep markdown files under `/llms/*.md` are currently Spanish-only (canonical marketing content). The English index above already surfaces every entity, feature and metric they contain. - [/llms-full.txt](https://sismos.es/llms-full.txt) — extended version with full text of the key pages (product, Campofrío case, capture guide, articles) in Spanish. - [/llms.txt](https://sismos.es/llms.txt) — Spanish index (mirror of this file). - [/llms/producto.md](https://sismos.es/llms/producto.md) — full Visor IA datasheet (ES). - [/llms/caso-carnico.md](https://sismos.es/llms/caso-carnico.md) — Campofrío success case with metrics (ES). - [/llms/captura-dataset.md](https://sismos.es/llms/captura-dataset.md) — full industrial dataset capture guide (ES). - [/llms/articulo-implantar-ia.md](https://sismos.es/llms/articulo-implantar-ia.md) — playbook to deploy AI on the shop floor (ES). - [/llms/articulo-eficiencia-operativa.md](https://sismos.es/llms/articulo-eficiencia-operativa.md) — impact of computer vision on OEE and trim losses (ES). - [/llms/calculadora-roi.md](https://sismos.es/llms/calculadora-roi.md) — methodology and formulas of the ROI calculator (ES). - [/llms/faq.md](https://sismos.es/llms/faq.md) — extended FAQ (ES). - [/llms/tienda.md](https://sismos.es/llms/tienda.md) — hardware catalog with pricing (ES). - [/llms/partners.md](https://sismos.es/llms/partners.md) — institutional and accelerator ecosystem (ES). - [/llms/integraciones.md](https://sismos.es/llms/integraciones.md) — PLC/SCADA, ERP/MES, camera and cloud integrations (ES). - [/llms/bventure.md](https://sismos.es/llms/bventure.md) — BVenture Challenge programme for early adopters (ES). ## Secondary commercial pages ### Shop — Hardware URL: https://sismos.es/tienda SISMOS hardware catalog: - **Visor IA Node** — Complete edge node with local AI processing (Hailo-8/Hailo-10H), 24V output to PLC, Modbus TCP/IP connection. From €1,700/unit. - **CSI-over-LAN camera module** — Raspberry Pi HQ sensor (IMX477, 4056×3040, rolling shutter) or Global Shutter (IMX296, 1456×1088), M12 8mm f/2.0 optics with 650nm IR-cut filter, reach up to 30m over Ethernet, no need for an industrial PC next to the camera. - No mandatory monthly fees in the base model; software license included. ### Partners and accelerators URL: https://sismos.es/partners Institutional ecosystem backing SISMOS: CaixaBank AgroBank (Destino4YFN), BEAZ (Diputación Foral de Bizkaia), Polo Positivo, Cámara de Comercio de Burgos, Digis3, LIN3S, AWS Startups, DIHBU, SolidWorks, Fundación Caja de Burgos, SPRI (Basque Government). ### Technical integrations URL: https://sismos.es/integraciones Visor IA native integrations with industrial and enterprise systems: - **PLC / SCADA**: Modbus TCP/IP (canonical registers V1.1), OPC-UA, 24V digital signals, Profinet (on the roadmap). - **ERP / MES**: REST API with token authentication, event webhooks (defect detected, alert, maintenance), CSV/JSON export. - **Cameras**: CSI (Raspberry Pi HQ IMX477 and Global Shutter IMX296) — productive; USB UVC — staged; RTSP, GigE Vision, proprietary SDKs (Basler/IDS) — lab. - **Cloud platforms**: AWS (SageMaker for training), Supabase (customer backend), model export to YOLO/COCO/Pascal VOC formats for use outside SISMOS. ### BVenture Challenge URL: https://sismos.es/bventure Commercial challenge for early adopters with special pilot deployment terms. Only active events are publicly visible. See also /eventos/:slug for occasional sector events (e.g. /eventos/bventure-bilbao-2026).