Tool cognition for robots

June issue now live

Tee, K.P., Cheong, S., Li, J. et al. A framework for tool cognition in robots without prior tool learning or observation.

  • Keng Peng Tee
  • Samuel Cheong
  • Gowrishankar Ganesh

Nature Machine Intelligence is a Transformative Journal; authors can publish using the traditional publishing route OR via immediate gold Open Access.

Our Open Access option complies with funder and institutional requirements.


  • An end-to-end machine learning approach that can learn which mechanisms determine cell fate and competition from a large time-lapse microscopy dataset is developed. The approach makes use of a probabilistic autoencoder to learn an interpretable representation of the organization of cells, and provides cell fate predictions that can be tested in drug screening experiments.

    • Christopher J. Soelistyo
    • Giulia Vallardi
    • Alan R. Lowe
  • Deep learning methods can provide useful predictions for drug design, but their hyperparameters need to be carefully tweaked to give good performance on a specific problem or dataset. Li et al. present here a method that finds appropriate architectures and hyperparameters for a wide range of drug design tasks and can achieve good performance without human intervention.

    • Yuquan Li
    • Chang-Yu Hsieh
    • Xiaojun Yao
  • Exoskeletons can assist movement in upper limb impairments to recover mobility and independence, but rigid or heavy exoskeletons can be impractical. Georgarakis and colleagues have developed a soft, tendon-driven device that assists shoulder movements and counteracts gravity to reduce muscle fatigue.

    • Anna-Maria Georgarakis
    • Michele Xiloyannis
    • Robert Riener
  • Robots usually learn to use tools from direct experience or from observing the use of a tool. While knowledge can be transferred between similar tools, novel and creative use of tools is challenging. Tee and colleagues present a method where skill transfer does not come from experience of using other tools but from using the robot’s own limbs.

    • Keng Peng Tee
    • Samuel Cheong
    • Gowrishankar Ganesh
  • While reinforcement learning can be a powerful tool for complex design tasks such as molecular design, training can be slow when problems are either too hard or too easy, as little is learned in these cases. Jeff Guo and colleagues provide a curriculum learning extension to the REINVENT de novo molecular design framework that provides problems of increasing difficulty over epochs such that the training process is more efficient.

    • Jeff Guo
    • Vendy Fialková
    • Atanas Patronov
Nature Machine Intelligence aims to bring different fields together in the study, engineering and application of intelligent machines. We publish research on a large variety of topics in machine learning, robotics, cognitive science and a range of AI approaches. We also provide a platform for comments and reviews to discuss emerging inter-disciplinary themes as well as the significant impact that machine intelligence has on other fields in science and on society.
Publishing online monthly from January 2019, Nature Machine Intelligence is interested in the best research from across the fields of artificial intelligence, machine learning and robotics. All editorial decisions are made by a team of full-time professional editors.
Nature Machine Intelligence is run by a team of full-time editors. The Chief Editor is Liesbeth Venema who was previously a physics editor at Nature. Trenton Jerde started in March 2018, Yann Sweeney joined in July and Jacob Huth joined in November 2018, completing the team.
Nature Machine Intelligence publishes original research as Articles. We also publish a range of other content types including Reviews, Perspectives, Comments, Correspondences, News & Views and Feature articles.
Contact information for editorial staff, submissions, the press office, institutional access and advertising at Nature Machine Intelligence