AI-empowered Robotics solutions for the Logistics Industry

Speaker:        Dr. Xiaoyu Ge
                Dorabot Inc. and Australian National University (ANU)

Title:          "AI-empowered Robotics solutions for the Logistics
                 Industry"

Date:           Monday, 11 March 2019

Time:           4:00pm - 5:00pm

Venue:          Lecture Theater F (near lift 25/26), HKUST

Abstract:

We are living in an era where robotics and artificial intelligence begin
to shape the future of various industries ranging from agriculture to
transportation, among which the logistics industry has become one of the
first areas radically transformed by AI and robotics techniques.

Since the emergence of AI, many problems in logistics domains have been
formulated and investigated as AI problems such as decision problems (bin
packing), planning problems (assembling, routing), and combinatorial
optimisation problems (task allocation, scheduling). Meanwhile, a
multitude of robots have been deployed in various logistics scenarios to
automate processes that involve tedious and repetitive tasks.  One
promising robotic solution is an automated warehouse that operates a large
group of robots without human interventions.

Hence, one natural question to ask is how AI techniques could help to
design an automatic warehouse and to optimise the performance of a large
system of robots created to perform different tasks, and what capabilities
these individual robots should have to accomplish routine tasks.

In this talk, I will first introduce a typical scenario in logistics and
our robotics solutions to that scenario.  Then I will present several
topics that I believe are significant for both academic research and for
industry applications. The topics will cover perception, multiagent task
and motion planning, combinatorial optimisation with spatial and physical
constraints, and the necessity of having explainable AI in our robotic
systems. For each of the topics, I will explain why the topic is important
and our methodology in approaching the topic.


**********************
Biography:

Dr. Xiaoyu Ge is the lead algorithm developer at Dorabot and a honorary
lecturer in the college of computer science at the Australian National
University (ANU). Prior to joining Dorabot, he worked as a postdoc fellow
in ANU. He received his Ph.D. degree in Artificial Intelligence from ANU
in 2017, his bachelor's degree in computer science with first class
honours from ANU and his bachelor's degree in information technology from
Monash University. His main research interests are in spatial
representation and reasoning, physical reasoning, and task planning in
robotics. He has been the key co-organiser of the Angry Birds AI
competition for seven years. He served as PC member of IJCAI19, IJCAI18
and KR18 and reviewer of TCIAIG.