Day 1 :
Keynote Forum
Ashitey Trebi-Ollennu
NASA Jet Propulsion Laboratory, USA
Keynote: The InSight Mission: Instrument Deployment Robotic System
Time : 10:00-10:45
Biography:
Ashitey Trebi-Ollennu is the Product Delivery Manager for the InSight Mars Mission Instrument Deployment System, Instrument Deployment System operations
Team Chief and a technical Group Lead in the Robotic Manipulation & Sampling Group at NASA Jet Propulsion Laboratory, California Institute of Technology, where
he has been since 1999. He has received his PhD degree in Control Systems from Royal Military College of Science, Cranfi eld University, UK in 1996 and BEng
from Queen Mary College, University of London, UK in 1991. He was a Research Scholar at Institute of Complex Engineered Systems, Carnegie Mellon University
from 1997 to 1999. He is a recipient of over a dozen NASA Group Achievement Awards. He is a Fellow of the Institution of Engineering and Technology, UK and a
Fellow of the Royal Aeronautical Society, UK, Senior Member IEEE RAS and IEEE SMC. He is also a Fellow of the Ghana Academy of Arts and Sciences and has
severed as a Guest Editor of the IEEE Robotics and Automation Society Magazine, Special Issue on Mars Exploration Rovers (June 2006). He has been on PhD
committees at the Robotic Institute of Carnegie Mellon University, Pittsburgh, USA and is the Founder of the Ghana Robotics Academy Foundation.
Abstract:
The InSight (Interior Exploration using Seismic Investigations, Geodesy and Heat Transport) Mars Lander Mission science
goal is to understand the formation and evolution of terrestrial planets through investigation of the interior structure and
processes of Mars. Th e science payload for InSight consist of a seismometer, heat fl ow probe and a precision tracking system to
measure the size and state of the core, mantle and crust. Additional payload includes a meteorology package, a magnetometer,
two color cameras and a 4 DOF robotic with a grapple at the end-eff ector. Th e InSight Instrument Deployment System (IDS)
is comprised of the Instrument Deployment Arm (IDA), arm-mounted Instrument Deployment Camera (IDC), landermounted
Instrument Context Camera (ICC) and control soft ware. Th e IDS is responsible for precision instrument placement
(seismometer and heat fl ow probe) on a planetary surface that will enable scientist to perform the fi rst comprehensive surfacebased
geophysical investigation of Mars. IDA has 1.9 m reach with four degrees of freedom: Yaw (shoulder azimuth) and three
pitch joints (shoulder elevation, elbow and wrist). Each joint has a temperature sensor and heater with a dust seal to prevent
contamination of the motor and gearbox. IDC allows visual confi rmation of deployment steps, as well as acquisition of the
stereo image pairs used to create a 3D map of the workspace. IDC also provides engineering images of solar arrays, payload
deck and instruments. ICC provides context images and redundant worksite imagery. Th is talk will give a brief overview of
InSight IDS and present test results and operations challenges for this fi rst of kind instrument placement and installation on
the Martian surface.
Biography:
Lin Zhou has obtained his PhD degree from the University of Glasgow, UK. He is currently the Program Director of Cognitive Education at Watson Education of IBM.
He leads the R&D of intelligent tutor and cognitive cloud fabric by developing and applying artifi cial intelligence capabilities. He has broad patents and publications
to his credit.
Abstract:
121 million children are out of school worldwide. 40% college students do not graduate aft er six years. Th e US student debt
has risen to over 1.3 trillion dollars. Access, outcome and cost are the key challenges to the society which have ignited
rapid transformation in the education industry. Th is article reports the latest development in the application of Artifi cial
Intelligence (AI) to education solutions in 3 dimensions: Content enrichment, student insight and personalized learning
delivery. Machining learning is shift ing traditional digitized learning content towards self-descriptive learning packages. Th ey
are learner aware, commerce aware and standard aware. In the student information dimension, AI is enabling educators to
gain greater insight into student learning style, predict performance risk and prescribe personalized remediations. Natural
Language Processing (NLP) analyzes student dialog with intelligent tutoring solutions to assess students’ competency mastery
and matches it with pedagogical models to deliver personalized learning at anytime, anywhere and judgment free. Th e exciting
new learning science development in the application of machining learning to study the brain wave (EEG) in a virtual reality
environment is also discussed
Keynote Forum
Timothy Sands
Naval Post Graduate School, USA
Keynote: Automated Mechanical Control
Time : 11:55-12:40
Biography:
Timothy Sands has completed his PhD at the Naval Postgraduate School and Postdoctoral studies at Stanford University and Columbia University, USA. He is an
Associate Dean of the Naval Postgraduate School’s Graduate School of Engineering and Applied Science. He has previously served as a Chief Academic Offi cer,
Dean and Senior Military Professor of the Air Force Institute of Technology. He has published research prolifi cally in archival journals, conference proceedings, book
chapter and holds one patent in spacecraft attitude control.
Abstract:
Automated control courses tend to strongly emphasize feedback control while minimizing or neglecting treatment of feedforward
controls which leads to systems that contain inherent lagging controllers. Furthermore, neglecting feed-forward
controls restrict engineers from taking advantage of adaptive feed-forward techniques oft en adapt control commands based
upon errors tracking trajectories and/or estimation errors. Direct adaptive control techniques typically directly adapt the
control signal without translation of estimated parameters. Indirect adaptive control techniques indirectly adapt the control
signal by translating the estimates of unknown system parameters to formulate a control signal. Th e adaptation rule is derived
using a proof that demonstrates the elimination of tracking errors (the true objective) and demonstrates stability, which is
complicated by the nonlinear closed loop system. Th is presentation will elaborate on such techniques applied to rotational
mechanics with time-varying mass
Keynote Forum
Bogdan Gabrys
Bournemouth University, UK
Keynote: Automatic composition, optimisation and adaptation of multicomponent predictive systems
Time : 12:40-13:25
Biography:
Bogdan Gabrys is a Data Scientist and Chair in Computational Intelligence at the Faculty of Science and Technology, Bournemouth University, UK. After many
years of working at different Universities, he moved to the Bournemouth University in January 2003 where he has founded and acted as a Head of Data Science
Institute, a Director of the SMART Technology Research Centre and Head of the Computational Intelligence Research Group within the School of Design, Engineering
& Computing and the Faculty of Science and Technology. His research, consulting and advisory activities have concentrated on the areas of data science,
complex adaptive systems, computational intelligence, machine learning, predictive analytics and their diverse applications. In particular, he has pursued the
development of various statistical, machine learning, nature inspired and hybrid intelligent techniques especially targeting data and information fusion, learning
and adaptive methods, multiple classifi er and prediction systems, processing and modeling of uncertainty in pattern recognition, diagnostic analysis and decision
support systems. He is an accomplished author with over 150 publications, frequently invited speaker at international events and sought after data science expert.
Abstract:
We are currently experiencing an incredible, explosive growth in digital content and information. According to IDC, the digital
universe in 2020 will be 50 times as big as in 2010 and that it will double every two years. Research in traditionally qualitative
disciplines is fundamentally changing due to the availability of such vast amounts of data. In fact, data-intensive computing has
been named as the fourth paradigm of scientifi c discovery and is expected to be a key in unifying the theoretical, experimental
and simulation based approaches to science. Th e commercial world has also been transformed by a focus on big data with
companies competing on analytics. Data has become a commodity and in recent years has been referred to as the new oil. Th ere
has been a lot of work done on the subject of intelligent data analysis, data mining and predictive modeling over the last 50 years
with notable improvements which have been possible with both the advancements of the computing equipment as well as with
the improvement of the algorithms. However, even in the case of the static, non-changing over time data there are still many
hard challenges to be solved which are related to the massive amounts, high dimensionality, sparseness or inhomogeneous
nature of the data to name just a few. What is also very challenging in today’s applications is the non-stationarity of the data
which oft en change very quickly posing a set of new problems related to the need for robust adaptation and learning over time.
In scenarios like these, many of the existing, oft en very powerful, methods are completely inadequate as they are simply not
adaptive and require a lot of maintenance attention from highly skilled experts, in turn reducing their areas of applicability.
In order to address these challenging issues and following various inspirations coming from biology coupled with current
engineering practices, we proposed a major departure from the standard ways of building adaptive, intelligent predictive systems by utilizing the biological metaphors of redundant but complementary pathways, interconnected cyclic processes,
models that can be created as well as destroyed in easy way, batteries of sensors in form of pools of complementary approaches,
hierarchical organization of constantly optimized and adaptable components. In order to achieve such high level of adaptability
we have proposed novel fl exible architectures which encapsulate many of the principles and strategies observed in adaptable
biological systems. Th e proposed approaches have been extensively and very successfully tested by winning a number of
predictive modeling competitions and applying to a number of challenging real world problems including pollution/toxicity
prediction studies, building adaptable soft sensors in process industry in collaboration with Evonik Industries or forecasting
demand for airline tickets covering the results of one of our collaborative research projects with Luft hansa Systems. Following
the drive towards automation of predictive systems building, deployment and maintenance, recent work at Prof. Gabrys' group
resulted in an approach and an open-source soft ware which allows to automatically compose, optimize and adapt multicomponent
predictive systems (MCPS) potentially consisting of multiple data preprocessing, data transformation, feature and
predictive model selection and post-processing steps.