Department of Computer Science1
Trust Management in Autonomous Ecosystem
Danish Iqbal
danish@mail.muni.cz
2
What are we talking about?
Trust Management
Why We Need the Trust?
Autonomous System
Solution for Trust Building
Concept of Digital Twin
3
Introduction (Trust Definition)
Discipline Definition of Trust
Sociology Subjective probability that another party will perform an action that will
not hurt my interest under uncertainty and ignorance.
Trust in
Ecosystem
Trust is a subjective expectation an agent has about another's future
behavior based on the history of their encounters.
Trust in
Automation
Trust can be described as “the attitude that an agent will help
achieve an individual’s goal in a situation characterized by
uncertainty and vulnerability.
Psychology Cognitive learning process obtained from social experiences based on
the consequences of trusting behaviors.
4
Introduction (Autonomous ecosystem)
̶ A system using sensors, actuators, control units, and other physical objects to control and protect a physical
infrastructure.
̶ Autonomous ecosystem are the tight integration of and coordination between computation with physical processes.
̶ Ecosystem include embedded systems and networks to monitor and control physical processes.
̶ Future ecosystem will be much stronger in
̶ Adaptability, autonomy, efficiency, functionality, reliability, safety, and usability.
̶ The challenge is the interactions between the embedded computers and the physical processes.
̶ Need to understand the joint dynamics of HW, SW, networks and physical processes.
5
Why we need the trust?
6
Levels of Vehicle Automation
7
Technology Timeline
Multiple generations of technology will co-exist on our roads for many years.
8
Literature Timeline
9
Trust Research
10
Recent Highlights
̶ SAE J3016 has been formally validated by the US Department of Transport.
̶ Tesla Motors Inc., BMW, Ford Motor Co. and Volvo Cars have all promised to have fully
autonomous cars on the road within five years.
̶ Alphabet Inc.’s (Google) autonomous test vehicles will surpass 3.5 million test miles on
public roads by May 2019.
̶ China has set a goal for 10-20% of vehicles to be highly autonomous by 2025, and for 10%
of cars to be fully self-driving in 2030.
̶ Nvidia and Mercedes-Benz announced intention to develop “cognitive car” using embedded
AI technology.
11
Two Distinct Cultures
̶ Divide into two Major category:
INFORMATION INDUSTRY
Security Culture
Major challenge to create a unified culture for these two very different industries.
AUTOMOTIVE INDUSTRY
Safety culture
12
Research Problem (1/2)
̶ Modeling and building trust in autonomous ecosystem is a challenging pace for both the
industry and academia [1].
̶ It requires a knowledge of both backgrounds cyber and physical field. Computer, Networks,
sensors, and actuators are played a vital role in the cyber world and the physical environment
that continuously communicates with each other.
̶ The physical environment is continuously evolving, at runtime, and in ways that cannot be
completely anticipated at the design stage.
13
Research Problem (2/2)
̶ The US National Academy of Engineering [2] has accumulated 14 major challenges in
Engineering of the 21st century that one of them is the verification of the CPS system in the
physical environment at runtime.
̶ The statistic report of the National Highway Traffic Safety Administration stated that there is
more than 5 million car accident annually in the united states and more than 2 million injuries
or fatalities. These ratios will be increased in the coming days because of overpopulation [3].
14
Modeling, Design, Analysis: Proposed Methodology
̶ Modeling is the process of gaining a deeper understanding
of a system through imitation. Models specify what a system does.
̶ Design is the structured creation of artifacts.
̶ It specifies how a system does what it does. This includes optimization.
̶ Analysis is the process of gaining a deeper understanding of a system through dissection.
It specifies why a system does what it does (or fails to do what a model says it should do).
15
Proposed Solution
̶ Concept of Digital Twin and Trust Assessment (Run-time verification)
̶ NASA was the first to propose the concept of digital twins for the cyber-physical systems. The present objective is to
advance the ideas and practice of digital twins across various industries, from manufacturing to healthcare.
16