Abstract
Advances in contemporary 3D scanning and bespoke robotic technologies have enabled architectural structures to be directly constructed from naturally grown wood. However, design precedents using natural wood logs are still dominated by design approaches using predefined geometric models. The limit of this approach lies in the necessity to model the form of every structural member based on the captured geometries of all the materials before design begins. Moreover, human designed rules for joining irregular components are limited and solutions are prone to be limited by empirical knowledge. In this paper, we introduce a method for assembling natural wood log structures with higher goals autonomously using robotic stochastic assembly and deep learning. The novelty of this method is that the design of structures does not rely on prior-knowledge of the to-be-assembled materials but is generated by assembling materials iteratively. A vision system with a position suggestion network based on convolutional neural networks (CNNs) was implemented and trained to drive an industrial robotic arm for negotiating between the topological changes from potential connections and the local assembly constraints of the log. A robotic hand-eye coordination database recording the assembly of birch logs has been established and small-scale wood structures were built by the trained robot. Results show that our robot can find desired structural configurations autonomously and can assemble unfamiliar batches of wood logs. The cost and gain of using stochastic assembly and deep learning as a design strategy are discussed and future research on using different learning strategies and large-scale implementations are laid out.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
BionicCobot. https://www.festo.com/group/en/cms/12746.htm. Accessed 1 Jun 2018
Smooth-on. https://www.smooth-on.com/products/mold-star-20t/. Accessed 1 Jun 2018
Windows Kinect. https://developer.microsoft.com/en-us/windows/kinect. Accessed 1 Jun 2018
Processing. https://processing.org/. Accessed 1 Jun 2018
Grasshopper Meshedit. http://www.food4rhino.com/app/meshedit/. Accessed 1 Jun 2018
Taco v0.70. http://blickfeld7.com/architecture/rhino/grasshopper/Taco/. Accessed 1 Jun 2018
Python Software Foundation. https://www.python.org/. Accessed 1 Jun 2018
TensorFlow. https://www.tensorflow.org/. Accessed 1 Jun 2018
Eversmann, P., Gramazio, F., Kohler, M.: Robotic prefabrication of timber structures: towards automated large-scale spatial assembly. Constr. Robot. (2017). https://doi.org/10.1007/s41693-017-0006-2
Funkhouser, T., Shin, H., Toler-Franklin, C., Castañeda, A.G., Brown, B., Dobkin, D., Rusinkiewicz, S., Weyrich, T.: Learning how to match fresco fragments. J. Comput. Cult. Herit. (JOCCH) 4(2), 7 (2011)
Goodfellow, I., Bengio, Y., Courville, A., Bengio, Y.: Deep Learning, vol. 1. MIT Press, Cambridge (2016)
Johns, R.L., Foley, N.: Bandsawn bands. In: McGee, W., Ponce de Leon, M. (eds.) Robotic Fabrication in Architecture, Art and Design 2014, pp. 17–32. Springer International Publishing Switzerland (2014)
Landa, Y., Galkowski, D., Huang, Y.R., Joshi, A., Lee, C., Leung, K.K., Malla, G., Treanor, J., Voroninski, V., Bertozzi, A.L.: Robotic path planning and visibility with limited sensor data. In: Proceedings of the American Control Conference, ACC, pp. 5425-5430. Piscataway, NJ (2007)
Levine, S., Finn, C., Darrell, T., Abbeel, P.: End-to-end training of deep visuomotor policies. J. Mach. Learn. Res. 17(39), 1–40 (2016)
Levine, S., Pastor, P., Krizhevsky, A., Quillen, D.: Learning hand-eye coordination for robotic grasping with large-scale data collection. Int. J. Rob. Res. 37(45), 421–436 (2018)
Mnih, V., Kavukcuoglu, K., Silver, D., Graves, A., Antonoglou, I., Wierstra, D., Riedmiller, M.: Playing Atari with deep reinforcement learning. arXiv preprint arXiv:1312.5602 (2013)
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A.A., Veness, J., Bellemare, M.G., Graves, A., Riedmiller, M., Fidjeland, A.K., Ostrovski, G.: Human-level control through deep reinforcement learning. Nature 518(7540), 529–533 (2015)
Pinto, L., Gupta, A.: Supersizing self-supervision: Learning to grasp from 50 k tries and 700 robot hours. In: Proceedings of the IEEE International Conference on Robotics and Automation, ICRA 2016, pp. 3406-3413. Stockholm (2016)
Rutten, D.: Evolutionary principles applied to problem solving. In: AAG10 Conference, Lecture at the International Conference on Advances in Architectural Geometry (AAG) 2010, Vienna (2010).
Schindler, C., Tamke, M., Tabatabai, A., Bereuter, M., Yoshida, H.: Processing Branches: reactivating the performativity of natural wooden form with contemporary information technology. Int. J. Archit. Comput. 12(2), 101–115 (2014)
Self, M., Vercruysse, M.: Infinite variations, radical strategies. In: Sheil, B., Menges, A., Glynn, R., Skavara, M. (eds.) Fabricate: Rethinking Design and Construction, pp. 30–35. UCL Press, London (2017)
Srinivas, A., Jabri, A., Abbeel, P., Levine, S., Finn, C.: Universal planning networks. arXiv preprint arXiv:1804.00645 (2018)
Steinhagen, G., Braumann, J., Brüninghaus, J., Neuhaus, M., Brell-Çokcan, S., Kuhlenkötter, B.: Path planning for robotic artistic stone surface production. In: Reinhardt, D., Saunders, R., Burry, J. (eds.) Robotic Fabrication in Architecture, Art and Design 2016, pp. 122–135. Springer International Publishing Switzerland (2016)
Toler-Franklin, C., Brown, B., Weyrich, T., Funkhouser, T., Rusinkiewicz, S.: Multi-feature matching of fresco fragments. ACM Trans. Graph. (TOG) 29(6), 1–12 (2010)
Van Den Berg, J., Abbeel, P., Goldberg, K.: LQG-MP: optimized path planning for robots with motion uncertainty and imperfect state information. Int. J. Robot. Res. 30(7), 895–913 (2011)
Wang, Y.: Hooke park biomass boiler house. In: Menges, A., Schwinn, T., Krieg, O.D. (eds.) Advancing wood architecture: A computational approach, part 4, chapter 12. Routledge, London (2016)
Acknowledgements
We would like to specially thank Andy Zeng and Shuran Song from the Computer Science department at Princeton University for their inspiring discussions at the early stage of the research. We would also like to thank staff William Tansley and Grey Wartinger for their constant lab support and the research funding provided by the School of Architecture of Princeton University during summer.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Wu, K., Kilian, A. (2019). Designing Natural Wood Log Structures with Stochastic Assembly and Deep Learning. In: Willmann, J., Block, P., Hutter, M., Byrne, K., Schork, T. (eds) Robotic Fabrication in Architecture, Art and Design 2018. ROBARCH 2018. Springer, Cham. https://doi.org/10.1007/978-3-319-92294-2_2
Download citation
DOI: https://doi.org/10.1007/978-3-319-92294-2_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-92293-5
Online ISBN: 978-3-319-92294-2
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)