Skip to content
NEW: CREATE AN AI-ASSISTED TENDER IN 2 MINUTES — TRY IT FREE
Back to search

Develops AI-based vision systems for automated visual quality control in industrial manufacturing.

DE

wenglor Deevio GmbH

BerlinNew on Amplifa

HRB 202058 B, AG Berlin (Charlottenburg)Commercial Register
GmbHLegal Form
> 98 %AI classification accuracy (BASF project)
up to 70 images/sInspection frequency (ampoule inspection)

About deevio

wenglor Deevio GmbH (deevio) is a Berlin-based AI and image processing company that offers machine learning solutions for automated visual quality control in industrial manufacturing. The company develops customized vision systems with industrial cameras and deep learning software, as well as retrofit solutions that upgrade existing image processing systems and reduce their false rejects – according to the company, from up to 50 percent down to as little as 1 percent. The core product is the AI Box, a GPU-based mini-computer with PLC connection on which trained deep learning models for image classification, segmentation and object detection run in real time; classification takes as little as 0.004 seconds per image. The scope of services includes data acquisition, a labeling tool, model training by the data science team, system integration with partners (cameras, lighting, machine control, PLC), as well as data analysis and data export for root cause analysis in production. Customers come, among others, from the pharmaceutical, automotive, precision tool and food industries, including BASF Polyurethanes and Takeda. The company is a spin-off from the wattx incubator of the Viessmann Group and was acquired in 2023 by the wenglor sensoric group (Tettnang).

Site
Berlin
Processes
Automated visual quality control (AI vision systems, new build), Retrofit of existing image processing systems (reduction of false rejects), AI image classification (defect detection OK/NOK), AI segmentation (error localization at pixel level), AI object detection, Visual inspection of pharmaceutical ampoules (retrofit AVI system, up to 70 images/s), Visual inspection of pharmaceutical packaging (up to 30 images/s), 100% visual quality control of die-cast parts (engine parts, cycle time 50 seconds per part)
Materials
Magnesium, Steel, Iron, Metal (general, inspected components)
Industries
Pharmaceutical industry, Automotive, Precision tools, Food & Beverage, Packaging industry, Die casting
  1. 01

    Consulting & requirements analysis

  2. 02

    Feasibility study

  3. 03

    Proof of Concept (hardware & data acquisition on site)

  4. 04

    Data acquisition & labeling

  5. 05

    Training of the deep learning model

  6. 06

    Integration on the AI Box with PLC connection

  7. 07

    Acceptance (Factory Acceptance Test)

  8. 08

    Industrialization with system integration partners

Inspected Materials & Applications

Mg

Magnesium

Inspected die-cast components (engine parts) as part of 100% visual quality control.

Stahl

Steel

Inspected metal components, among others in the die-casting context.

Fe

Iron

Inspected metal components, among others in the die-casting context.

Location

Berlin

DE

Headquarters, development and data science

  • AI Box (GPU-based mini-computer with PLC connection)
  • Labeling tool
  • AI software toolkit (classification, segmentation, object detection)

Industries

Pharmaceutical industryAutomotivePrecision toolsFood & BeveragePackaging industryDie casting

Processes & technical limits

ProcessesToleranceMax. partWeightLot sizeMaterials
Automated visual quality control (AI vision systems, new build)Inspection–––––
Retrofit of existing image processing systems (reduction of false rejects)Inspection–––––
AI image classification (defect detection OK/NOK)Inspection–––––
AI segmentation (error localization at pixel level)Inspection–––––
AI object detectionInspection–––––
Visual inspection of pharmaceutical ampoules (retrofit AVI system, up to 70 images/s)Inspection–––––
Visual inspection of pharmaceutical packaging (up to 30 images/s)Inspection–––––
100% visual quality control of die-cast parts (engine parts, cycle time 50 seconds per part)Inspection––––Magnesium, Steel, Iron
Visual inspection of precision toolsInspection–––––
Retrofit of a 3D measuring system with AI model (BASF, accuracy > 98%, 0.004s per image)Inspection–––––
Engineering–––––

4 Anlagen & Maschinen

MachineManufacturer / modelQuantityWorking areaYear builtSite
AI Box (GPU-based mini-computer for AI inspection, with PLC connection, real-time control)deevio AI-Box–––Berlin
Industrial cameras / image acquisition setups (project-specific)–––––
3D measuring device (customer system, retrofit, BASF)––––Lemförde (BASF)
AVI system (automated visual inspection system, pharma, retrofit)–––––

Additional services

AI vision systems for visual quality controlRetrofit of existing image processing systemsAI Box (GPU hardware with PLC connection)AI software toolkit (classification, segmentation, object detection)Labeling toolActive LearningData analysis/data exportSystem integration with partners (cameras, machine control)

Reference customers

BASF Polyurethanes GmbHTakedaGerman automotive supplier (anonymized)

History

  1. Spin-off from the wattx incubator of the Viessmann Group

  2. 2019

    CEO Donato Montanari is re-elected to the board of the VDMA Machine Vision trade association

  3. 2023

    Acquisition by the wenglor sensoric group (Tettnang); Berlin location is retained, management passes to Christian Vollrath

Personalised demo

Ready to see Amplifa Marketplace in action?

From a structured request to the right manufacturing partner: AI-assisted matching, verified suppliers and transparent metrics — in one place for industrial companies across the DACH region.

Request a demo