Frequently Asked Questions

Clobotics is a Singapore-headquartered physical AI company founded in 2016. It digitizes real-world field operations using computer vision, robotics, and artificial intelligence — primarily for wind energy asset inspection and repair, and retail CPG execution intelligence. Clobotics has launched projects across 40+ countries worldwide.

Clobotics has offices in Singapore, Shanghai, Seattle (R&D), and Denmark — but our reach goes far beyond our headquarters. We have staff and field teams distributed across the globe, ready to support local project launches wherever you operate. For enterprise clients running cross-border programs, this matters: our team members bring diverse international backgrounds and a strong track record of executing projects across multiple countries simultaneously. Whether your wind portfolio spans Europe and APAC, or your retail brand operates across dozens of markets, Clobotics can mobilize locally while coordinating globally.

Clobotics currently serves two industries: wind energy (autonomous blade inspection, internal inspection, lightning protection testing, robotic leading-edge repair, and fleet asset intelligence via the IRIS platform) and retail & CPG (shelf intelligence, execution compliance, out-of-stock detection, share of shelf, branded asset tracking, and store layout analytics via RIaaS). Our computer vision and robotics capabilities are applicable across many other field-intensive industries — if you're curious about what's possible in your sector, we're always open to exploring new applications together.

Clobotics Wind Intelligence is a full-stack blade lifecycle platform connecting IBIS autonomous drone inspection, KIWI internal blade inspection, HUMMINGBIRD lightning protection testing, SPARROW robotic leading-edge repair, and IRIS asset intelligence — giving wind operators a single end-to-end workflow from inspection evidence to completed repairs.

Clobotics runs a five-stage AI pipeline on every inspection image: blade classification, blade segmentation, then defect recognition using sliding-window parallel analysis. The blade classification model achieves 99% accuracy, segmentation 98% accuracy, and defect recall exceeds 95% overall — with major-defect recall above 98% and major-defect precision above 80%. In practice, on a 400-image turbine inspection the AI automatically clears roughly 240 images as defect-free, reducing human review time from 1.5 hours per turbine to 25 minutes.

Accuracy in defect detection is a data problem as much as a technology problem. Clobotics has conducted blade inspections across 40+ countries — building one of the largest real-world blade defect datasets in the industry. Every completed inspection is archived into a data lake that feeds continuous model retraining on a PDCA loop. The result is a model that improves with every campaign, not one trained on a static dataset and left unchanged.

Most inspection vendors cover external blade surfaces only. Clobotics inspects the full blade: IBIS drones capture the exterior at sub-millimeter resolution (avg 0.66 mm/pixel, best case 0.25 mm/pixel); the KIWI robot crawls inside the blade for internal structural defect detection; and HUMMINGBIRD performs lightning protection resistance testing — all three data streams are cross-referenced on the IRIS platform to give a complete internal-external blade health picture. No other vendor offers all three in a single integrated workflow.

All Clobotics inspection systems are designed for single-person operation. IBIS (external blade inspection) and HUMMINGBIRD (lightning protection testing) each complete a full turbine in approximately 15 minutes — one operator, no crew at height. KIWI (internal blade inspection) takes 15–20 minutes per blade. SPARROW (robotic leading-edge repair) is the fastest robotic LEP solution available: from prep and setup to finished repair, a single blade offshore takes as little as 38 minutes.

SPARROW is Clobotics' fully autonomous leading edge protection (LEP) repair robot, deployed by heavy-lift drone. It requires no rope-access crew or jack-up vessel, completes a blade repair in under 1 hour onshore and 38 minutes per blade offshore, and can treat up to 6 blades per day. In 2023 SPARROW completed the world's first offshore robotic blade repair. The coating is polyurea-based with a lab-tested lifespan of 15 years.

IRIS is Clobotics' cloud-based asset intelligence platform that unifies external, internal, and LPS inspection data in one workspace. It provides defect annotation with precise location mapping, severity rating, automated reports, fleet-wide dashboards across turbines and sites, and digital blade records retained across ownership and service transfers — all with no installation required.

Clobotics RIaaS (Retail Intelligence as a Service) is a retail execution intelligence platform for CPG brands, bottlers, and retailers. It connects store image capture, computer vision recognition, KPI generation, and field execution workflows so teams can monitor shelf availability, pricing, promotions, branded assets, and planogram compliance from one system.

Clobotics can extract shelf availability, out-of-stock signals, share of shelf, linear space, planogram compliance, price tag accuracy, promotional display compliance, display type, branded asset placement, and selected freshness or date-based signals from standard store images. Depending on the project, outputs can include recognized products, dimensions, prices, POP materials, expiry or manufacturing dates, and execution KPIs at shelf, bay, row, or store level.

Yes. Clobotics is built for real retail conditions, including stacked products, beverage multipacks, and products shown at different angles. The platform can be configured to count all visible facings or destacked facings depending on the KPI definition, and it can treat multipacks as distinct sellable SKUs when that matches the commercial reality in a market.

Clobotics supports all three modes. Teams can run cloud-based real-time recognition, fully offline capture with delayed upload, or hybrid workflows that combine both inside the same audit task. For a typical image containing around 80 facings, server-side recognition usually completes in under 5 seconds excluding network upload and download time, which makes it practical for in-store execution workflows.

Clobotics provides both raw recognition outputs and business-ready KPI outputs. Customers can receive delivery through APIs, file services, web portals, or BI tools, with export formats including CSV, JSON, XML, and Parquet. This makes it possible to use Clobotics data inside existing dashboards, data lakes, and enterprise reporting environments.

Yes. Clobotics supports planogram and rule-based compliance checks at multiple levels, including shelf, bay, row, and column. Depending on the program, the rules engine can evaluate precise matching, statistical compliance logic, and score-based compliance outputs, so brands can adapt the workflow to both strict planograms and more flexible execution standards.

Clobotics combines customer-provided master data with field discovery workflows and AI-assisted product registration. The platform is designed to onboard new markets, detect previously unregistered products, and update models continuously as assortments evolve. In active projects, model updates are typically rolled out on a regular cadence so new products and edge cases are incorporated without waiting for a major relaunch.

Clobotics is built on a cloud-native architecture designed for large multi-market deployments. Today, the platform processes around 1 million images per day, with peak observed throughput around 100 images per second. The system is designed to scale across large SKU counts, multiple countries, and multiple categories without changing the field workflow for frontline teams.

Clobotics REA app supports iOS 11.0 or later and Android 9.0 or later. For customers that want tighter system integration, Clobotics provides iOS and Android SDKs, a RESTful OpenAPI platform, and integration patterns for enterprise systems such as SAP, Salesforce, file-based pipelines, and BI environments. Deeplink integration is also available for customers that need looser app-to-app coupling.

Customers retain ownership of their original data. Clobotics supports secure delivery and access through APIs, object storage, file services, and reporting layers, while also providing role-based access control and multi-tenant isolation for enterprise deployments. Operationally, Clobotics supports global rollouts with regional service coverage, multilingual app support, structured deployment and training programs, and 24/7 handling for priority technical issues.