Milan Zdravković presented on next-generation enterprise information systems (NG-EIS) and challenges in research opportunities. Key points include hypothesizing properties and a generic architecture for NG-EIS, such as omnipresence, multiple identities, and dynamic configurability. Implementation of EIS remains challenging due to high costs and failure rates. Semantic interoperability, requirements engineering, and model-based systems engineering are important areas of focus for overcoming IoT system challenges.
Next-generation Enterprise Information Systems - Research opportunities and challenges
1. NEXT-GENERATION EIS – RESEARCH
OPPORTUNITIES AND CHALLENGES
Milan Zdravković
Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
2. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
ABOUT
My research
Semantic interoperability/formal modelling, MBSE, IoT; domains:
SCM, orthopaedic devices
86 papers, 14 journal papers with IF, 308 citations, h=10,
My academic/community work
Guest edited 3 journal special issues, never said no to review/evaluation
assignment
Chairing ICIST conference series since 2014
Member of IFAC TC5.3, IFIP WG8.1
Fascinated with research policy
Hands-on
1 start-up, 2 spin-offs
Enjoy working with local IT communities
Can’t live without music, skiing and travel
http://www.masfak.ni.ac.rs/milan.zdravkovic/
3. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
IFAC TC 5.3 TECHNICAL COMMITTEE ON
ENTERPRISE INTEGRATION AND NETWORKING
TC 5.3 fosters research in
enterprise integration and interoperability
enterprise architectures,
enterprise engineering methods,
enterprise modelling
https://tc.ifac-control.org/5/3
4. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
FIND US @
13th OTM/IFAC/IFIP International Workshop on
Enterprise Integration, Interoperability and
Networking (EI2N 2018)
Valletta, Malta, October 24-25, 2018
16th IFAC Symposium on Information Control
Problems in Manufacturing
Bergamo, Italy, June 11-13, 2018
5. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
INTRODUCTION
English translation of Welsh: “I am
not in the office at the moment.
Please send any work to be
translated”
6. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
DATA
NEVER
SLEEPS
5.0
https://www.domo.com/learn/data-never-sleeps-
5?aid=ogsm072517_1&sf100871281=1
7. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
TODAY, DATA IS NOT CREATED ONLY BY PEOPLE
Pratt & Whitney’s Geared
Turbo Fan (GTF) engine is
fitted with 5,000 sensors
and can generate up to
10GB of data each second
Data generated by the
aerospace industry alone
could soon surpass the
magnitude of the
consumer Internet. http://aviationweek.com/connected-aerospace/internet-
aircraft-things-industry-set-be-transformed
11. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
THE FIRST BUZZWORD TODAY:
SENSING ENTERPRISE
“an enterprise anticipating future decisions by using multi-
dimensional information captured through physical and virtual
objects and providing added value information to enhance its
global context awareness” (FInES cluster, Santucci et al, 2012)
Key tech enablers for Sensing Enterprise are IoT and big data
12. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
WHAT ARE IOT PLATFORMS?
Cloud-based, Platform-as-a-Service
Typical features
connectivity as a service, monitoring and maintenance of devices
(including firmware updates), data visualization, data analytics, basic
application logics through alerts and triggers.
Core market niches (Zdravkovic et al, 2016)
Domain-specific (turnkey) platforms
Technology-specific platforms
M2M connectivity services
Full-scale generic IoT middlewares
Supporting services
13. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
BASIC INFRASTRUCTURE COMPONENTS OF
IOT PLATFORMS
IoT Gateway
IoT Platform
IP Device
MQTT
HTTP REST
Thin client
Platform
specific
client SDK
Monitoring
Data analytics
Maintenance
Data storage
External services
– not native IoT
platforms
Data storage
API
Business rules,
e.g. IFTTT
API
Advanced
analytics
API
Advanced
visualization
API
….
API
IP Endpoints
Custom, domain
specific backend
API
Device-oriented services
Security
Data visualization
Data-oriented services
HTTP REST
PAN Endpoints
ZigBee
Bluetooth
6LowPAN
Platform
specific
gateway
SDK
Client
Data pre-processing
Scalability
Infrastructural services
CoAP
PAN Device
Trigger-response-
alert rules
App-oriented services
Scripting engine -
interpreter
Reasoning
14. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
WHAT’S WRONG WITH IOT PLATFORMS?
IoT platforms are typically driven by models of the trivial complexity; they
support very simple data structures and almost no business logic
implementation
Application Logic (AL) processing in IoT platforms is a bottleneck
AL not expressive, if-then rules, or..
=> NOT EFFECTIVE
AL complex - hidden in code, this code is written by using specific platform
SDKs.
=> NOT FLEXIBLE
AL not formal, no semantics used, no interoperable with other platforms/systems
=> NOT INTEROPERABLE
IoT systems are today usually managed centrally
15. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
HYPOTHESES: PROPERTIES AND GENERIC,
ABSTRACT ARCHITECTURE OF NG-EIS
Omnipresence
Multiple identities
Model Driven
Architecture
Awareness/inclusive
sensing
Openness Dynamic configurability
Computational flexibility
Distributed identities
Models infrastructure Services infrastructure
Deployment platform
Interoperability infrastructure
Core execution environment
Security
Trust
Scalability
Performance
Next Generation Enterprise Information System
16. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
http://www.tamingdata.com/2010/07/08/the-project-management-tree-swing-cartoon-past-and-present/
17. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
OH YES, EIS IMPLEMENTATION IS STILL A
PROBLEM
In 2010, the mean Enterprise Resource Planning (ERP) systems’
implementation cost was $5.48 million, and the average implementation
time-frame was 14.3 months (Galy and Sauceda, 2014)
On average, large IT projects run 45% over budget and 7% over time,
while delivering 56% less value than predicted (McKinsey, 2012)
85% of existing devices worldwide are based on unconnected legacy
systems (IDC, 2013)
19. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
INTEROPERABILITY FOR IOT
Interoperability is the ability to
communicate with peer systems and
access the functionality of the peer
systems (Vernadat, 1996)
In IoT:
ReST (Representational State Transfer)
CoAP (Constrained Application
Protocol)
MQTT (Message Queue Telemetry
Transport)
XMPP (Extensible Messaging and
Presence Protocol)
Google Cloud Internet of Things Core, End-to-End Example
https://cloud.google.com/iot/docs/samples/end-to-end-sample
20. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
INTEROPERABILITY AT SEMANTIC LEVEL
60-80% of the resources in data sharing projects are spent on addressing the
issues of semantic heterogeneities (Doan et al, 2004)
IoT services, namely devices’ capabilities can be semantically represented, so
the process of discovery and selection is done by using SPARQL queries
(Song et al, 2010)
Scalability, flexibility and other issues cannot be resolved with centralized
approach.
Each of the devices must be represented by (and connected to) its virtual
counterpart, which is implemented as autonomous software agent (Katanasov et al,
2008).
IoT platform is then only a run-time environment for these software agents.
22. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
BUILDING BLOCKS FOR SEMANTIC
INTEROPERABILITY IN IOT
In order to interoperate with another device, each device must sense, perceive, interpret and
understand data, sent from another device and act (operate) upon this understanding.
Abstract the heterogeneity of devices, so one device can better understand the capabilities of
another
Sensor Network (SSN) Ontology (Compton et al, 2012)
23. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
INTEROPERATION ENGINE
NOT a middleware, it does not mediate traffic, but meaning
Interoperating Engine is basically a formal reasoner with extensions
Computed values can be defined in OWL (Iannone and Rector, 2008)
Alternative: pre-processing (with different modeling approach) which will classify data in ranges of interest.
PR-OWL - Probabilistic extensions, based on Bayesian network (Costa and Laskey, 2006)
Probabilistic/fuzzy can be defined by Fuzzy OWL 2 (Bobillo and Straccia, 2010), there are
reasoners for that (DeLorean)
25. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
REQUIREMENTS ENGINEERING FOR IOT
SYSTEMS
To be transformed to models – models make the decisions of RE
explicit
Key concepts for RE for IoT (Zambonelli, 2016)
goals - desirable situations or state of the affairs that should be activated
and which should be decomposed to system requirements
identification of stakeholders and users who will manage or use systems’
functionalities and from which functional requirements should be elicited
26. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
CHALLENGES FOR RE AND MANAGEMENT
Elicitation of non-functional requirements, such as security and
privacy, performance, reliability, energy usage, impact to
environment
Prioritization
New requirements will emerge, related to decision-making
capabilities and higher level of automation
Incoming data as a source for new requirements
New change management strategies
27. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
MODEL-BASED SYSTEM ENGINEERING (MBSE)
FOR IOT SYSTEMS
OMG Standards for IoT
System Modeling Language (SysML - INCOSE and OMG)
language for systems specification, analysis, design, verification and validation
Less implicit formalisms are used at application design level - Domain Specific Languages (DSL)
Conceptual models as starting effort towards DSL (Patel and Cassou, 2014)
Midgar IoT platform with DSL (Garcia et al, 2014)
ThingML for forward engineering of IoT (Harrand et al, 2016)
To facilitate interaction modeling in a heterogeneous environment, many researches rely on the explicit -
formal models
Devices will be formally represented and managed, registered, aligned, composed and queried through suitable
abstraction technologies (Kotis and Katasonov, 2013)
Formal models are convenient for integrating different views – levels of abstraction
Formal models can be used also for system and application design
28. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
CHALLENGES FOR MBSE
Efficiency related
Cost (of time consumed and errors made) of manual creation of models,
including maintaining dependencies between different model views
Model mapping and transformation
Rationality related
Cost of choices made in very heterogeneous environment of non-synced
models, languages, beliefs and preferences of the system engineers.
Lack of tools for validating model consistency, completeness and dependability
Lack of formalisms for representing business logics
29. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
FORMAL FRAMEWORK FOR MULTI-FACETED
VIEW TO IOT ECOSYSTEM
Model validation problem could
be solved by using Formal
Specification Techniques (FST)
Use of formal models and
associated methods (semantic
annotation, ontology matching)
Towards ontology-driven IoT
systems, which use formal
models in a runtime
30. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
SYSTEM ARCHITECTURE FOR IOT
IoT system is inherently distributed
Agent-based node in IoT systems is independent, self-
governing software and hardware integrated system. It is
capable to sense, to interpret the sensation, make the best
informed decision on that interpretation and finally, act
upon it
intelligent goal-based agent
Composition approaches
Centralized
Decentralized
31. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
CHALLENGES RELATED TO SA
Trust models, which will ensure that proper decisions are made
in environment of uncertain credibility, high reliability and
security concerns
Self-management capability of IoT systems, by realizing the
context-awareness and self-adaptation properties of the
individual agents (Ayala et al, 2015)
32. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
MATURITY MODELS FOR IOT
MM used to continuously assess the capability of a system
to maintain its operation at the agreed quality levels,
to evolve, sometimes even in very short time frames, on demand
MM typically include a sequence of levels (or stages) that form an
anticipated, desired, or logical path from an initial state to maturity
(Becker et al, 2009)
IoT MMs
MM and tool to assess manufacturing companies adherence (Schumacher et al, 2016)
or readiness (Lichtblau et al, 2015) (Rockwell Automation, 2014) to the Industry 4.0
vision
MM to enable organizations to assess their readiness for IoT and to compare
themselves against others with IoT initiatives (Halper, 2016) or progress in
implementing IoT (Vachterytė, 2016)
SELF-AWARE
SENSING
SMART
ACTIVE
33. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
CHALLENGES RELATED TO MM
Maturity assessment reference ontology, which generically and
formally defines improvement paths and maturity levels
In context, defined by the domain ontologies
With implicit problem representations, defined by application
ontologies
35. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
IOT POLICY AND
REGULATORY
ASPECTS
Those issues affect
each of the
elements of the IoT
systems
implementation
framework
36. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
EXTENDED DOMAIN FRAMEWORK FOR IOT SYSTEMS
IMPLEMENTATION
37. Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
CASE STUDY – IOT ENABLES MANUFACTURING
SHOP-FLOOR
38. NEXT-GENERATION EIS – RESEARCH OPPORTUNITIES AND
CHALLENGES
Milan Zdravković
Baltic DB&IS 2018, July 1-4, Trakai, Lithuania
THANK YOU FOR YOUR ATTENTION
Q&A