Folie 1

1
Collaborative Research Centre 637
2

• CRC Collaborative Research Centre
• Funding instrument of the German Research
  Foundation (DFG),
• Funds from the Federal Ministry of Education and
  Research and from the Federal States of Germany,
• Long-term university research centre (3 x 4 years
  12 years),
• Cross-disciplinary research programme,
• Aims to create a core research focus at a
  university.

3
Source Bundesamt für Güterverkehr, 2002

• Room for Improvement in Truck Transport
• 23 of covered distances are deadhead,
• only 77 truck utilisation in terms of distance,
• only 67 truck utilisation in terms of loading
  weight (even 58 in long-distance traffic).
• continuous traffic increase,
• limited infrastructure.

4
(No Transcript)
5
Change Drivers of Logistic Processes
6

• hierarchical IT structure
• global information processing
• central control

• distributed IT structure
• with global communikation
• local information processing
• autonomous and decentralized control

7

• intelligent cargo, transit equipment and
  transportation systems
• permanent localisation, identification and
  communication with and between these logistic
  objects
• autonomous cooperating
• logistic objects

8
(No Transcript)
9

• Project Domain A
• Modelling Foundations for Autonomous Logistics
  Processes
• Research into and development of theoretical
  foundations for autonomous systems
• Investigation of the limits of conventional and
  autonomous control
• Modelling of autonomous logistic processes

10
Logistics System
Task Layers within the CRC 637
11
Project Domain A A1 Fundamental Studies
Windt A2 Sustainable Management
Müller-Christ A3 Monitoring of Autonomous
Systems Hülsmann A4 Rule-based Graph
Transformation Kreowski A5 Dynamics of
Autonomous Systems Scholz-Reiter / Wirth
Project Domain B B1 Reactive Planning and
Control Scholz-Reiter / Görg B2 Adaptive
Business Processes Modelling and
Methodology Scholz-Reiter B3 Mobile
Communication Networks and Models Görg B4
Knowledge Management Herzog B5 Risk Management
Herzog / Schumacher B6 Sensor Systems Lang /
Laur B7 Autonomous Adaptation of Vehicle
Schedules Kopfer
Project Domain Z Z1 CRC Management Herzog Z2
Demonstrator Scholz-Reiter
12
(No Transcript)
13

• Autonomy (Hülsmann, Windt)
• Definition frame for Autonomy
• CRC-wide, consistent understanding of the
  paradigm Autonomy
• Scenarios Modelling (Görg, Kreowski, Windt)
• Discussion of different models and modelling
  approaches
• Modelling of autonomy
• System Platform (Herzog)
• CRC-wide, agent-based software platform
• Software basis for prototypical implementations
• Applications (Scholz-Reiter, Kopfer)
• Identification of potential application fields

14
Involved Faculties and Institutes

• Faculty 1 Physics / Electrical Engineering
• Research Group Communication Networks (Görg)
• Institute for Microsensors, -actuators and
  -systems (Lang)
• Institute for Electromagnetic Theory and
  Microelectronics (Laur)
• Faculty 3 Mathematics / Computer Science
• Research Group Mathematical Systems Theory
  (Wirth)
• Research Group Artificial Intelligence (Herzog)
• Research Group Theoretical Computer Science
  (Kreowski)
• Faculty 4 Production Engineering and Technology
• Research Group Planning and Control of Production
  Systems (Scholz-Reiter, Windt)
• Research Group IT-based Applications in
  Production Engineering (Schumacher)
• Faculty 7 Business Studies and Economics
• Research Group Logistics (Kopfer)
• Research Group Sustainable Management
  (Müller-Christ)
• Research Group Management of Sustainable
  Corporate Development (Hülsmann)

15
(No Transcript)
16

• The Collaborative Research Centre 637
• Autonomous Cooperating Logistic Processes -
• A Paradigm Shift and its Limitations
• focuses on a long-term, interdisciplinary field
  of research,
• combines the investigation of highly dynamic
  logistic processes with the development and
  application of recent IC technologies,
• is based on a well proven interdisciplinary
  combination of the necessary research groups at
  the University of Bremen,
• has significant relevance for practical
  applications and high potential for future
  technology transfer,
• has a structural effect for the university and
  the region of Bremen.

17
Working definition of autonomy Autonomy
describes processes of decentralized decision
making in networking structures. It requires
interacting objects in nondeterministic systems
that are able to make autonomous decisions. The
aims of autonomy are improved robustness and
positive emergence of the whole system by
distributed and flexible handling of the dynamics
and complexity.
18
Logistics System
Task Layers within the CRC 637
19
(No Transcript)
20
products
raw materials
21
depot
customers
22
depot
customers
23

• modelling and simulation of autonomous material
  flow
• analysis of the dynamics of logistics processes
• development of new planning and control methods
• benchmarking
• prototypical implementation (demonstrator)
• practical applications (transfer projekts)

24
Logistics System
Task Layers within the CRC 637
25
goals and conditions for the transport
26
(No Transcript)
27

• goals
• very large and scalable simulations
• support of the demonstrator
• technological basis multi-agent systems
• flexibility, robustness, and scalability
• modelling of autonomy
• tools
• MAS platform JADE
• knowledge representation by ontologies (OWL)

28
Logistics System
Task Layers within the CRC 637
29
A4
Modelling of Autonomous Logistic Processeswith
Rule-based Graph Transformation
Autonomous UnitsSyntax,Semantics, and Case
Study Ludo
Karsten Hölscher, Renate Klempien-Hinrichs,
Peter Knirsch, Hans-Jörg Kreowski, Sabine Kuske
Theoretical Computer Science, TZI, Department of
Mathematics and Computer Science International
Graduate School for Dynamics in Logistics
30
A4
Goals

• Approach for the Modelling of
• Autonomous Logistic Processes with
• Rule-based Graph Transformation
• semantical analysis of the generative power
• relation to the theory of concurrency
• modelling concept for the visualization of the
  semantical processes
• systematic verification approach with respect to
  Model Checking, Constraint Solver and Inductive
  Theorem Prover
• embedding and comparison of other process models

31
A4
Concepts
32
A4
Syntactic schema
community of autonomous units
overall goal (if there is any), terminal
environments
set of autonomous units
Com (Aut, Init, Goal)
both specified by graph class expressions
initial environments
autonomous unit
aut (goal, rules, control)
individual
33
A4
Examples
die
rules
start
turnroll
roll-again
34
A4
Examples
35
A4
Semantic requisites
graph transformation approach

• environments G set of graphs

• set X of of environment class expressions
  with admitted environments SEM(g) ? G for g ?
  X

• set C of control conditions with control
  relation SEM(c) ? G ?G for c ? C

changes of environment CHANGE ? G ?G
36
A4
Sequential semantics of autonomous units
sequential process of aut (goal, rules,
control)
with dynamic environment given by CHANGE G0,
G1, G2, with (for all i) Gi ? Gi1 and
(Gi, Gi1) ? SEM(control) for some r ? rules or
(Gi, Gi1) ? CHANGE
r
SEQCHANGE(aut) set of all sequential processes of
aut
reachability of goal Gj ? SEM(goal)
37
A4
Sequential semantics (global)
sequential process of Com (Aut, Init, Goal)
SEQ (Com) set of all sequential processes of Com
reachability of overall goal Gj ? SEM(Goal)
Observation SEQ(Com) SEQCHANGE(aut) for aut ?
Aut if CHANGESEQ(Com-aut)
38
A4
Ludo environment graph
Ludo (player(b),player(y), player(g),player(r),
die, initconf, finish)
39
A4
Ludo game situation
40
A4
Ludo game situation
41
A4
Process semantics

• sequential
• Set of parallel processes
• concurrent
• starting in initial environments,
• composed of rule applications of the
  participating units,
• with every unit autonomously controlling its
  actions.

identifying those processes in which the overall
goal or individual goals are achieved.
42
A4
Relations
process models in computer science
petri nets
multiagent systems (cf. Wooldridge) cooperation
with B4
community ofautonomous units
swarm intelligence cellular automata
graph transformation units
board and card games
scenariosfrom AK Szenarien
logistic process models cooperation with A1
43
A4
Petri nets
Idea every transition as autonomous unit
Place/Transition system (N,m0)
C(N,m0)
S/T-2-CAU
firing
ruleapplication
firing sequences
SEM(C(N,m0))
Theorem 1-1-relation between firing sequences
with seq./par. firing and seq./par.
processes 1-1-relation between processes in
Condition/Event systems and concurrent processes
44
A4
Advantages

• formal with rigorous operational semantics
• visual (based on graphs)
• supporting verification
• generalizing existing approaches
• admitting comparisons with other approaches