Daniel Schäle
Daniel Schäle has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
Erik Kyrkjebø
HVL Robotics, Norway
2 shared publications- Erik Kyrkjebø
Western Norway University of Applied Sciences
1 shared publication Martin F. Stoelen
HVL Robotics, Norway
1 shared publication- Martin F. Stoelen
Western Norway University of Applied Sciences
1 shared publication Martin Fodstad Stølen
1 shared publication
Latest Publications
- Open Access
- 6 May 2025
- Published by Frontiers Media SA in Frontiers in Robotics and AI
- Vol. 12, 1569476
Learning computer-aided manufacturing from demonstration: a case study with probabilistic movement primitives in robot wood carving
Computer-Aided Manufacturing (CAM) tools are a key component in many digital fabrication workflows, translating digital designs into machine instructions to manufacture physical objects. However, conventional CAM tools are tailored for standard manufacturing processes such as milling, turning or laser cutting, and can therefore be a limiting factor - especially for craftspeople and makers who want to employ non-standard, craft-like operations. Formalizing the tacit knowledge behind such operations to incorporate it in new CAM-routines is inherently difficult and often not feasible for the ad hoc incorporation of custom manufacturing operations in a digital fabrication workflow. In this paper, we address this gap by exploring the integration of Learning from Demonstration (LfD) into digital fabrication workflows, allowing makers to establish new manufacturing operations by providing manual demonstrations. To this end, we perform a case study on robot wood carving with hand tools, in which we integrate probabilistic movement primitives (ProMPs) into Rhino’s Grasshopper environment to achieve basic CAM-like functionality. Human demonstrations of different wood carving cuts are recorded via kinesthetic teaching and modeled by a mixture of ProMPs to capture correlations between the toolpath parameters. The ProMP model is then exposed in Grasshopper, where it functions as a translator from drawing input to toolpath output. With our pipeline, makers can create simplified 2D drawings of their carving patterns with common CAD tools and then seamlessly generate skill-informed 6 degree-of-freedom carving toolpaths from them, all in the same familiar CAD environment. We demonstrate our pipeline on multiple wood carving applications and discuss its limitations, including the need for iterative toolpath adjustments to address inaccuracies. Our findings illustrate the potential of LfD in augmenting CAM tools for specialized and highly customized manufacturing tasks. At the same time, the question of how to best represent carving skills for flexible and generalizable toolpath generation remains open and requires further investigation.
- 6 May 2025
- Published by Frontiers Media SA in Frontiers in Robotics and AI
- Vol. 12, 1569476
Open AccessLearning computer-aided manufacturing from demonstration: a case study with probabilistic movement primitives in robot wood carving
Computer-Aided Manufacturing (CAM) tools are a key component in many digital fabrication workflows, translating digital designs into machine instructions to manufacture physical objects. However, conventional CAM tools are tailored for standard manufacturing processes such as milling, turning or laser cutting, and can therefore be a limiting factor - especially for craftspeople and makers who want to employ non-standard, craft-like operations. Formalizing the tacit knowledge behind such operations to incorporate it in new CAM-routines is inherently difficult and often not feasible for the ad hoc incorporation of custom manufacturing operations in a digital fabrication workflow. In this paper, we address this gap by exploring the integration of Learning from Demonstration (LfD) into digital fabrication workflows, allowing makers to establish new manufacturing operations by providing manual demonstrations. To this end, we perform a case study on robot wood carving with hand tools, in which we integrate probabilistic movement primitives (ProMPs) into Rhino’s Grasshopper environment to achieve basic CAM-like functionality. Human demonstrations of different wood carving cuts are recorded via kinesthetic teaching and modeled by a mixture of ProMPs to capture correlations between the toolpath parameters. The ProMP model is then exposed in Grasshopper, where it functions as a translator from drawing input to toolpath output. With our pipeline, makers can create simplified 2D drawings of their carving patterns with common CAD tools and then seamlessly generate skill-informed 6 degree-of-freedom carving toolpaths from them, all in the same familiar CAD environment. We demonstrate our pipeline on multiple wood carving applications and discuss its limitations, including the need for iterative toolpath adjustments to address inaccuracies. Our findings illustrate the potential of LfD in augmenting CAM tools for specialized and highly customized manufacturing tasks. At the same time, the question of how to best represent carving skills for flexible and generalizable toolpath generation remains open and requires further investigation.
- Open Access
- 1 January 2023
- Published by Norwegian Society of Automatic Control in Modeling, Identification and Control: A Norwegian Research Bulletin
- Vol. 44 (4) , 141-154
Programming Fine Manufacturing Tasks on Collaborative Robots: A Case Study on Industrial Gluing
This paper explores programming of fine manufacturing tasks using collaborative robots. We conduct a case study based on an industrial gluing task, comparing two programming approaches: Learning from Demonstration (LfD) and Computer-Aided Manufacturing (CAM). We investigate the suitability of these approaches for ad-hoc automation of fine manufacturing tasks by expert operators, and discuss the strengths and weaknesses associated with their usage. The case study reveals that there are benefits and limitations to both approaches. The CAM-based approach provides a precise path for execution without the need for robot programming expertise, but is strongly dependent on the quality of the gauging process. The LfD approach is intuitive and quick to set up, but is strongly dependent on the quality of the demonstrations. Our findings suggest that there is a potential for a hybrid solution combining the best of both approaches in a unified interface, and provide a foundation for future research on hybrid programming interfaces for fine manufacturing tasks using collaborative robots.
- 1 January 2023
- Published by Norwegian Society of Automatic Control in Modeling, Identification and Control: A Norwegian Research Bulletin
- Vol. 44 (4) , 141-154
Open AccessProgramming Fine Manufacturing Tasks on Collaborative Robots: A Case Study on Industrial Gluing
This paper explores programming of fine manufacturing tasks using collaborative robots. We conduct a case study based on an industrial gluing task, comparing two programming approaches: Learning from Demonstration (LfD) and Computer-Aided Manufacturing (CAM). We investigate the suitability of these approaches for ad-hoc automation of fine manufacturing tasks by expert operators, and discuss the strengths and weaknesses associated with their usage. The case study reveals that there are benefits and limitations to both approaches. The CAM-based approach provides a precise path for execution without the need for robot programming expertise, but is strongly dependent on the quality of the gauging process. The LfD approach is intuitive and quick to set up, but is strongly dependent on the quality of the demonstrations. Our findings suggest that there is a potential for a hybrid solution combining the best of both approaches in a unified interface, and provide a foundation for future research on hybrid programming interfaces for fine manufacturing tasks using collaborative robots.
- 31 October 2021
- Published by Springer Nature
- p. 273-283
Priors Inspired by Speed-Accuracy Trade-Offs for Incremental Learning of Probabilistic Movement Primitives
- 31 October 2021
- Published by Springer Nature
- p. 273-283
Priors Inspired by Speed-Accuracy Trade-Offs for Incremental Learning of Probabilistic Movement Primitives