TG relation: Yes, Perception
Areas mainly addressed: Agile Production, Robotics and AI/Big Data Value, B2B and innovation
Main questions to be answered:
· Synthetic training data generation requires model data for simulation. Do you have this data available and how exact is it?
· For unknown objects, real data is required in addition to synthetic data. Which ratio of real data to synthetic data is feasible in your use case?
· Which level of expertise regarding 3d vision and machine requirements are available in your area? (none, beginner, moderate, expert)
WS Description: Perception is one of the key technologies for enabling flexible production, such as pick and place, machine tending, assembly, and quality testing. 3 major trends can be observed. GOOD DATA, NOT BIG DATA: Synthetic data generation based on model data and enriched by real data create realistic ground-truth training data sets for machine learning. PLUG-AND-PRODUCE: The perception component of an automation solution must be easily adjustable to changing requirements. EASE-OF-USE: With robot vision expertise being a scarce resource, usability for robot users with little to now vision knowledge is a game-changer. These trends increase the flexibility making the RoI much easier to demonstrate, especially for SMEs. The data requirements for industrial use cases are elaborated in an interactive session with the attendees guided by three key questions. The goal is to enable flexible production.
Intended outcome: Workshops engaging non-roboticists to understand needs/goals (policy, commercial, technical), Workshop discussion topics of common interest, success stories, use cases, etc, Workshop on recent developments in technology or applications
Approach: 11:05 Introduction and definition of key statements/questions, Dr. Michael Suppa, CEO, Roboception GmbH
11:15 Towards Detecting and Grasping Transparent Objects, Prof. Markus Vincze, Professor, TU Vienna, Austria
11:25 Ocado Bot data orchestration’s, Dr. Radhita Gudipati. Ocado Technology
11:35 Data generation for Lab Automation,Dr. Patrick Courtney, CEO, Tec-connection, UK
11:45 Good Data for pick-and-place in Agile Production, Michael Suppa. CEO, Roboception GmbH, Germany
11:55 Interactive poll session/round table discussion with the audience
12:20 Conclusion and take home messages
Contributors: Dr. Michael Suppa, CEO, Roboception GmbH, Germany, confirmed
Prof. Markus Vincze, Professor, TU Vienna, Austria, confirmed
Dr. Patrick Courtney, CEO, Tec-connection, UK, confirmed
Dr. Radhika Gudipati, Research Coordinator (Robotics & AI), Ocado Technology, UK, confirmed
Further information: The workshop will be published on the company website (https://roboception.com/de/innovation-de/#workshop) and announced via Linkedin. The slides will be made available as pdf on the website.
Organisers:
· Michael Suppa, michael.suppa@roboception.de, Roboception GmbH
Event Timeslots (1)
Room K2 – Koncerthuset
-
15 March - Agile Production
Michael Suppa