Oct 16, 2026 – Oct 18, 2026 · flow1986, Offenburg

SMART AUTOMATION

Name: SMART AUTOMATION
Start: 2026-10-16T13:00:00Z
End: 2026-10-18T14:00:00Z
Location: flow1986, Offenburg

Join the Discord server

presented by BadenCampus

Results & photos

The final ranking, the winners and the photos from the event.

Results Photo album

About

black forest hackathon

The Black Forest Hackathon is the meeting place for visionaries who want to spend a weekend working with the latest technologies in a unique environment and atmosphere. We bring you together with like-minded people so that you can implement innovative projects from scratch in just 48 hours.

smart automation edition

At our SMART AUTOMATION hackathon, you’ll tackle real-world industrial challenges in smart production—from AI-powered robotics to cutting-edge Industry 4.0 solutions. With hands-on support from experienced automation experts, tech founders, and industry professionals, you’ll turn ideas into prototypes with real impact on manufacturing. Along the way, you’ll become part of the smartXautomation ecosystem, connecting with leading industrial companies, innovative startups, and potential partners while helping shape the path toward the autonomous factory.

Overview

Dates: Friday, October 16, 2026 at 3:00 PM – Sunday, October 18, 2026 at 4:00 PM
Venue: flow1986, Offenburg (DE)

Challenges · published only

BCT Technology AG

Drawing2CAD

Many companies have old design drawings for which no digital 3D CAD model is available. Often, only technical 2D drawings are available as PDF files, scans, or images. If these components need to be modified, replicated, or reused, design engineers must manually extract dimensions and geometry from the drawing and recreate the 3D model. This process is time-consuming, error-prone, and ties up valuable resources in the design department.

KOCH Freiburg GmbH

Codebound: Sandbox for AI Agents in AnythingMCP

AI agents currently face inefficiencies when processing large datasets, as every intermediate result must travel back through the model. In production, 1 in 8 calls returns over 50 KB, with the largest being 44 MB, making multi-step tasks slow, costly, or manual. This challenge focuses on enabling AI agents to execute small scripts directly where the data resides, within a secure sandbox environment in AnythingMCP, to streamline workflows and reduce latency.

Herrenknecht AG

Automated extraction and standardisation of drilling data for a global geodatabase

Herrenknecht requires reliable information about the subsurface for the development of tunnel boring machines. Borehole data from tunnel and mining projects around the world is available in various formats, languages and representations, and has so far had to be analysed manually. The challenge lies in automatically extracting and standardising this data, and recording it in a georeferenced form within a central, searchable database featuring an interactive world map.

VEGA Grieshaber KG

Tactile Intelligence: Object Localization with a Humanoid Robot Hand

Humans can pick up an object and identify its position and orientation just by touch. They can then place it at a specific location without seeing it. For a humanoid robot, this is much more difficult. The robot must use data from tactile sensors in its hand to understand where the object is and how it is oriented. Available Data & Resources: - Dexterity Hand RH56DFTP-2R/2L - 17 tactile sensors per hand - Python API for robot control and sensor data processing - Real-time tactile sensor data - Sample objects and technical documentation

The full brief of a challenge opens for registered participants.