Knowledge Engineering for Autonomous Agents


Jan. 2022 - Our paper on developing a hybrid architecture that enhances Deep Learning with size and spatial awareness has been accepted for publication at AAAI Spring Symposium on Machine Learning and Knowledge Engineering for Hybrid Intelligence (AAAI-MAKE).

Sep. 2021 - Prof. Enrico Motta gave the inaugural keynote titled "Combining Deep Learning and Common-Sense Reasoning in Visually Intelligent Robots" at Semantics 2021! - Our team joined forces with AIRLab (Politecnico di Milano) and won the "shopping cart" episode at the 2021 Smart Cities and Robotics Challenge!

June 2021 - Agnese presented an overview of her doctoral work at the European Semantic Web Conference (ESWC) PhD Symposium.

Mar. 2021 - We presented our work at AAAI-MAKE. In this paper we present a novel hybrid architecture (DL-based + knowledge-based) for object recognition where the typical size of objects is considered.

Oct. 2020 - Our latest work on integrating knowledge-based reasoning about object sizes within ML is now available on ArXiv. Be sure to visit our Github for the related code and dataset.

Sep. 2020 - We presented our work at KR2020! Watch the video here.


HanS is a Health and Safety Robot Inspector , which was developed by Intelligent Systems and Data Science (ISDS) robotics team.
HanS is an experiment in service robotics and has been designed to be able to patrol the Knowledge Media Institute (KMi) to try and identify possible violations to Health and Safety rules.
For instance, typical rules check that all fire extinguishers are in place and that emergency exits are free of obstacles and clearly labelled through the correct signs.

Currently, HanS is able to recognise many objects commonly found in office spaces automatically (without relying on ARTags), after being trained on only a few shots captured from the target environment.

HanS is aware of the typical size and spatial realtions of objects. HanS Knowledge Base was autonomously constructed and summarise from a number of large-scale resources, including WordNet, ShapeNet, Visual Genome, and Quasimodo.

Thanks its commonsense reasoning capabilities, HanS can autonomously assess the risks posed by the situations it observes.


Gianluca Bardaros Photo Gianluca BardaroResearch Associate
Agnese Chiattis Photo Agnese ChiattiPhD Research Student
Enrico Dagas Photo Enrico DagaSenior Project Officer
Enrico Mottas Photo Enrico MottaProfessor of Knowledge Technologies