Periskope provides a computer‑vision platform that uses deep‑learning models to automatically detect defects and monitor processes on manufacturing lines in real time. It ingests video from existing cameras, runs inference on edge or cloud GPUs, and delivers alerts and quality metrics through OPC‑UA, REST and MQTT APIs for seamless integration with MES, SCADA and ERP systems. The system supports transfer‑learning for rapid model adaptation, continuous drift monitoring, and can be deployed on‑premise, in the cloud, or in hybrid environments with end‑to‑end encryption.
Funding
$1.2M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
Manual visual inspection on production lines is labor‑intensive, subject to human variability, and cannot keep pace with high‑throughput manufacturing, leading to missed defects and inconsistent product quality. Additionally, integrating disparate camera systems into a unified monitoring workflow is complex and costly.
Solution
Periskope delivers a computer‑vision platform that automates defect detection and process monitoring using deep‑learning models trained on manufacturing data. Video streams from existing line cameras are ingested and processed in real time on edge or cloud GPUs, generating instant defect alerts and quality metrics. A model‑training toolkit supports transfer learning, enabling rapid adaptation to new product families with minimal labeled data. The platform exposes OPC‑UA and REST APIs for direct integration with MES, SCADA, and ERP systems, allowing automated corrective actions. All data are encrypted in transit and at rest, and the solution can be deployed on‑premise or in hybrid cloud environments to meet security and latency requirements. Continuous model performance monitoring detects drift and triggers automated retraining to maintain detection accuracy over time.
Target Audience
Primary customers are high‑volume manufacturers—automotive, electronics, consumer goods, and process industries—who need reliable, automated quality inspection. The solution is used by quality‑control engineers, production supervisors, and automation integrators responsible for line efficiency and product compliance.
Features
- Real‑time inference engine optimized for GPU and edge accelerators, handling up to 120 fps per camera stream
- Transfer‑learning pipeline with pre‑trained defect models that can be fine‑tuned on custom datasets in hours
- Multi‑camera synchronization and 3‑D reconstruction for inspection of complex geometries
- OPC‑UA, REST, and MQTT connectors for seamless integration with MES, SCADA, and ERP systems
- Automated KPI dashboards with statistical process control charts and defect trend analytics
- On‑premise, cloud, or hybrid deployment options with end‑to‑end AES‑256 encryption
- Continuous model drift detection and scheduled retraining workflows to sustain accuracy