Resource search

Search DataDrivenAEC

Open the full search page →

oculai

Crane-mounted AI cameras that turn construction sites into an automatic target/actual schedule

German ConTech company that mounts plug-and-play cameras and mobile-communication modules on tower/top-slewing cranes to film the entire construction site from a bird's-eye view. Computer vision and deep learning models trained on the footage detect and classify manual construction activities against a component and process taxonomy, automatically comparing planned vs. actual schedule progress and generating daily site reports, working-hour records, and productivity (effort-value) metrics.

€2.5M seed round (closed March 2023), led by High-Tech Gründerfonds (HTGF, its 700th investment) with Bayern Kapital, Axeleo Capital, the Leonard accelerator (VINCI), and angel investors including Arno Zinke (ex-Autodesk/Hexagon), Christoph Jentzsch (ex-Slock.it), and Tobias Ortmaier (ex-Yuanda Robotics) — per HTGF and Leonard/VINCI announcements. oculai was a member of Leonard's SEED 2022 cohort before closing the round. No further funding round has been publicly announced since March 2023. raisedEarly Stage
Founded 2021 · 20+ construction firms referenced on oculai's site (including HOCHTIEF, Implenia, Marti Holding, Zech Bau, Riedel Bau, GS Schenk); "Under The Hard Hat" reported four of Germany's five largest construction companies use oculai (unverified beyond that secondary source). ~20-22 employees per HTGF and RocketReach.
Website →← All Tools

Why This Tool Exists

The Problem

Site managers on large shell-construction (Rohbau) projects manually track progress, productivity, and daily documentation by walking the site and writing reports by hand — a process that is time-consuming, inconsistent, and often surfaces schedule delays and productivity shortfalls weeks after they occur, once rework or claims are already costly.

The Solution

oculai mounts AI-enabled cameras on site cranes to continuously film the whole jobsite, then uses computer vision to automatically detect and classify construction activities (tied to a component and process from the schedule), producing a live target/actual schedule comparison, auto-generated daily construction reports, and per-component labor/productivity (effort-value) metrics — without manual data entry.

How You Use It

Delivery Method
SaaS (WebApp)Hardware (crane-mounted camera + connectivity module)
Integrations
Primavera P6MS ProjectAsta Powerproject
Disciplines
Project Phases
Construction Administration
Project Types
Commercial, Infrastructure, Residential (multi-family/large-scale)

Data Transparency

Exactly what this tool uses and how

Input
What it needs
Required:Live video feed from crane-mounted cameras, Project schedule (target dates by component/process)
Optional:Material/component quantities (entered manually to generate effort-value/productivity metrics)
Formats:Primavera P6 file, MS Project file, Asta Powerproject file, Video stream (camera hardware, format not publicly specified)
Output
What you get
Format:WebApp dashboard with schedule, reports, and metrics; auto-generated daily construction report
Fields:Target vs. actual schedule status per component/process, Detected activity type and duration, Working hours and crew size per process/component, Short video clips of each detected activity, Weather data, Disruption/impediment log (for claims documentation), Effort-value (productivity) metrics per component, Projected completion date
Algorithm
How it works
Model:Proprietary deep learning computer-vision models (architecture not publicly disclosed); founder's academic background is in "Human Action Recognition with AI" (FAU Erlangen-Nuremberg research)
Accuracy:Not publicly specified — oculai and third-party coverage cite business-outcome claims (e.g., up to 25% shorter project durations, ~15% construction-management cost savings, projects "18% faster than average") rather than model-level detection accuracy/precision figures
Privacy
How your data is protected
Retention:Not publicly specified for project/video data; oculai's general website privacy policy states contact-form data is retained until deletion is requested or its purpose no longer applies. Per oculai's standard terms (AGB), customers get a 6-week window to retrieve their data in a standard format after contract termination.
Training:oculai's terms of service state it "may use anonymized customer data to monitor performance and improve software"; optional AI features may transfer project data to EU-based external subcontractors. Whether footage/imagery is used to train oculai's core detection models is not publicly specified.
Compliance:GDPR-aligned data processing (Art. 6, Art. 21 cited in privacy policy; EU data controller), No SOC 2, ISO 27001, or other third-party security certification publicly disclosed
API
Integration
Endpoint:Contact vendor for API access — no public API documentation found
Method:Not publicly documented

Use Cases

  • ·Automated target vs. actual schedule tracking on shell-construction (Rohbau) projects
  • ·AI-generated daily construction site reports (activities, weather, working hours)
  • ·Productivity/effort-value benchmarking per trade and component across projects
  • ·Early delay detection via 24/7 remote site monitoring from the project office
  • ·Disruption and impediment documentation to support claims

Pricing

Free
Not offered — no free tier; contract requires camera hardware provisioning
Pro
Monthly usage fee covering software access and camera provision, billed in advance; exact rates not publicly disclosed (per oculai's published AGB/terms). Initial contract term is 6 months, then auto-renews monthly. Installation, training, and data-migration services are billed separately per oculai's price list.
Enterprise
Custom pricing for multi-project/portfolio deployments — contact vendor
Sources & Research NotesResearch date: 2026-07-17
Fields Checked (16)
problem, solution, deliveryMethod, integrations, disciplines, projectPhases, vendor, website, fundingTotal, yearFounded, algorithm-approach, input-formats, output-fields, useCases, customerBase, pricing-model-structure
Not Found (7)
api-documentation, soc2-or-iso-certification, exact-data-retention-duration-for-video, algorithm-accuracy-methodology, exact-pricing-figures, ai-training-data-policy-for-video-footage, hosting-provider
About this research: This record summarizes the cited sources and fields checked. A research date does not establish hands-on testing. Consult the original sources for current details and limitations.