HiP-HOPS: A novel method and tool for dependability analysis and optimisation of systems
Submitting Institution
University of HullUnit of Assessment
Computer Science and InformaticsSummary Impact Type
TechnologicalResearch Subject Area(s)
Information and Computing Sciences: Artificial Intelligence and Image Processing, Computation Theory and Mathematics, Computer Software
Summary of the impact
The University of Hull has pioneered a novel method and tool for
dependability analysis and optimisation of critical engineering systems
known as Hierarchically Performed Hazard Origin and Propagation Studies
(HiP-HOPS).
a) HiP-HOPS (http:\\hip-hops.eu) has been successfully commercialised in
conjunction with software houses ITI GmbH (Germany) and ALL4TEC (France).
Over 30 licences for the tool have been sold since 2011 with total income
for all partners estimated at £300,000.
b) The method and tool have been taken up by large organisations
including Volvo, Toyota, Honda, Fiat, Continental, Germanischer Lloyd,
Embraer and Honeywell.
c) HiP-HOPS has contributed to the specification of EAST-ADL, an emerging
design language developed as an automotive industry standard, confirming
industrial reach and acceptability.
d) The Dependable Systems research group is pursuing impact on the new
automotive safety standard ISO-262626 and have contributed to setting up
the new IFAC DCDS workshop a key forum for disseminating research on
dependability to industry.
Underpinning research
The evaluation and optimisation of dependability of computer-based safety
critical systems is a very challenging problem. Over the last fifteen
years, work on model-based dependability analysis has resulted in new
approaches that partly automate and simplify the synthesis of
dependability evaluation models. HiP-HOPS is one of the first and
presently the most advanced among modern compositional dependability
analysis techniques. It was originally conceived by Professor Papadopoulos
during his PhD research at York and has been developed in Hull since 2001.
The technique has produced significant innovations and key findings within
the period all of which progress the state-of-the-art and have been
extensively published. These are listed below and mostly covered in
references mentioned in [1-3] and further detailed in the 8 submitted REF
outputs (Papadopoulos, Parker and Walker) which stem from new developments
in HiP-HOPS:
a) Fast linear-time algorithms for model-based synthesis of (temporal)
fault trees and (sequential, multiple failure mode) Failure Modes and
Effects Analyses (FMEAs).
b) A language for the description of inheritable and reusable component
failure patterns.
c) A new temporal logic (PANDORA) that facilitates analysis of temporal
fault trees.
d) Model-based algorithms for real-time diagnosis and correction of
failures.
e) Algorithms for design space exploration and multi-objective
optimisation of system designs via automatic model transformations using
meta-heuristics.
f) Contributions to the EAST-ADL error model annex and algorithms for
automatic allocation of safety requirements in the form of Safety
Integrity Levels.
A number of other contemporary dependability analysis techniques
including FSAP-NuSMV, Software Deviation Analysis, and DCCA have used
variants of state-modelling, model-checking and fault simulation as a
means of inferring the effects of component failures in a system. In
theory, these techniques offer a higher degree of automation than
HiP-HOPS. However, extra automation comes at a price that of much higher
computational complexity. The analysis of individual failure modes via
simulation or model-checking is computationally expensive and the often
forward, inductive nature of the analysis creates difficulties, especially
when combinations of failures need to be considered [2].
In HiP-HOPS, the analysis of propagation of failures is done by a
deductive algorithm which links effects on system outputs to causes in the
architecture. Synthesis of fault trees is achieved in linear time, and
although the subsequent fault tree analysis can still be time-consuming,
overall simplicity has enabled not only application of the technique to
large systems but also its unique combination with computationally greedy
heuristics such as Genetic Algorithms [1]. An additional difficulty with
most formal dependability analyses techniques is that they typically
define their own language for nominal and failure modelling which is not
always fully compatible with widely used design languages and tools.
HiP-HOPS, on the other hand, focuses only on failure modelling and can
easily complement design languages focusing on descriptions of nominal
behaviour. It has so far been demonstrated to work as an add-on to
EAST-ADL, Matlab Simulink, and Simulation X [3].
In summary, HiP-HOPS offers certain advantages to other contemporary
techniques, and these have driven the industrial impact of the method.
These advantages are: compatibility with a range of modelling notations;
scalability of the analysis; and unique capabilities for fault modelling,
temporal analysis and architectural optimisation via automatic model
transformations.
HiP-HOPS was developed in a string of European Projects [4-6] with input
from large industrial organisations across the transport industries who
have since become commercialisation partners and users of this research.
Papadopoulos (Senior Lecturer, 2001-2007; Reader, 2007-2011; Professor,
2011-present) has led the development of this work. Bottaci (Senior
Lecturer, 1986-present) has contributed to the formalisation of aspects of
HiP-HOPS Walker and Parker (both Research Assistants, 2008-2011;
Lecturers, 2011-present) are authors of the PANDORA temporal logic and
optimisation algorithms respectively. All the above mentioned staff have
been in the University of Hull for the duration of the REF period, and
further back in the period since 2001 when much of the underpinning
research that led to this impact case was carried out.
References to the research
Publications (all peer reviewed in high quality journals)
1. I Wolforth, M Walker, L Grunske, Y Papadopoulos (2010), Generalisable
Safety Annotations for Specification of Failure Patterns, Software
Practice and Experience, 40(5):453-483, DOI: 10.1002/spe.966
2. M Adachi, Y Papadopoulos, S Sharvia, D Parker, T Tohdo (2011), An
approach to optimization of fault tolerant architectures using HiP-HOPS,
Software Practice and Experience, 41:1303-1327, DOI: 10.1002/spe.1044
3. M Walker, M-O Reiser, S Tucci, Y Papadopoulos, H Lonn, D Parker, D-J
Chen (2013) Automatic Optimisation of System Architectures using EAST-ADL,
Journal of Systems & Software, 86(10): 2467-2487, DOI:
10.1016/j.jss.2013.04.001.
Grants (all peer reviewed, FP6-FP7 European grants)
1. (Duration: 2010 - 2013), MAENAD (FP7 Grant 260057) Modelling Analysis
Evaluation of Novel Architectures for Dependable Electric Vehicles (with
Fiat, Volvo, 4S, MetaH Continental, Delphi, French Atomic Authority, TU
Berlin, RIT Stockholm) European Commission, Principal Investigator :
Papadopoulos, £239,437
2. (Duration: 2008 - 2010), ATESST2 (FP7 Grant 224442) — Advancing
Traffic Efficiency and Safety through Software Technology (with Volvo
Technology, Volkswagen, Continental, French Atomic Authority, TU Berlin)
European Commission, Principal Investigator : Papadopoulos, £178,873
3. (Duration 2005 - 2009, partly in period), SAFEDOR (FP6 IP Grant
516278) — Safe Design Operation and Regulation, FP6 Integrated Project (53
partners including all major ship classification authorities in Europe,
shipyards, equipment manufacturers and operators) European Commission,
Principal Investigator : Papadopoulos, £213,078 (€20 million in total)
Details of the impact
HiP-HOPS has achieved industrial, economic and societal impact and it is
widely recognised as one of the techniques that define the
state-of-the-art in dependability analysis and optimisation of systems.
This section first presents the three strategic activities through which
impact was pursued in the period and then details the industrial, economic
and societal impact achieved.
Impact Strategy and Activities (2008-present):
a) Commercialisation of the HiP-HOPS tool and global engagement with
large industrial users, including Volvo, Fiat and Continental in Europe,
Denso and Toyota in Japan, Embraer in Brazil and Honeywell in the USA.
Over 30 licences of the tool have been sold to these organisations since
2011 (see details of impact including value of sales under heading
"Details of Impact Achieved" below and sources to corroborate impact:
3,4,5).
b) Technology transfer to industry through European Projects including
three projects in the automotive and shipping domains (see refs [4-6] in
section 3) and consultancies. Research has been transferred to R&D
departments and system design teams. A consultancy for the Flemish
Mechatronic Institute (2012) has led to seven HiP-HOPS licences deployed
in various companies that are members of this organisation (see details of
impact including benefits for companies involved under heading "Details of
Impact Achieved" and sources to corroborate impact: 1,2,6,7,8,9).
c) Organisation of events focused on HiP-HOPS and related model-based
dependability analysis technologies. Papadopoulos has co-organised and
co-chaired multi-session tracks and tool sessions in successive IFAC
symposia on Information Problems in Manufacturing and IFAC World Congress
in 2008. This activity led to the establishment of the IFAC Workshop on
Dependable Control of Discrete Event Systems which provides the main forum
for technology transfer to industry on the field of dependability. IFAC
DCDS'13 was co-organised by the University of Hull and co-chaired by
Papadopoulos (http://dcds13.net.dcs.hull.ac.uk/).
DCDS'13 was financially supported by Bosch and was attended by senior
representatives of British industries. There were three sessions on
state-of-the-art in Model-based Safety Assessment with strong
representation of HiP-HOPS. There was strong interest from the Head of the
Software Centre of Excellence in Rolls Royce, a seminar was given
(04/11/2013) and technology transfer to the company is being explored.
Details of Impact Achieved:
1. Economic impact (2008-present)
A consortium agreement with ITI GmbH (a German Computer-Aided Engineering
specialist software house), and Germanischer Lloyd was signed in May 2010,
and HiP-HOPS was commercially launched in January 2011. Germanischer Lloyd
(GL) is a major international register of shipping with enormous influence
in the industry and actively promotes this technology in the shipping
industry. This development has sprung out of SAFEDOR, the largest ever
project on safety funded by the EU with 53 partners and a total budget of
€20 million, where the University of Hull had a key role as provider of
technologies for automated safety analysis and design optimisation.
ALL4TEC in France, a company specialising in provision of tools and
services for system safety analysis, is also commercialising HiP-HOPS to
its clients in the context of a sub-licensing agreement with the
University of Hull.
Commercialisation activities have started to contribute directly to the
economy of the UK since 2011 via HiP-HOPS sales currently at £60,000 by
the University of Hull which controls the HiP-HOPS licence. Additional
benefits to commercialisation partners are estimated at £240,000. The
latter are arising from stake in the sale of HiP-HOPS, additional sales of
modelling tools linking to HiP-HOPS, and services such as training and
application modelling. There are also economic benefits via improvements
in the processes of safety analyses achieved by HiP-HOPS — which
contribute to reduced costs arising both from automation and efficiencies
but also safer product/system development for the industrial users of the
tool. It is difficult to quantify the latter, and no studies have been
done on this, but the growing uptake of the technology is very promising
in this respect, and included is a list of users that can provide
estimates on such benefits in this section.
2. Impact on industrial practice (2008 — present)
In a string of European and industrial funded projects, HiP-HOPS has
directly contributed to the specification of the error modelling
capabilities of EAST-ADL, an emerging architecture description language
developed as an automotive industry standard for the design of vehicle
control systems by a consortium of automotive companies. As a result of
this work, in 2011, the EAST-ADL association was founded to manage the
standardisation and evolution of the language, and promote its adoption in
the automotive industry. The University of Hull is a founding member of
this association and Papadopoulos is among two academics sitting on the
5-member board. This development is important because it confirms wide
reach and acceptability of the method and is preparing the ground for
further industrial and societal impact in the future.
HiP-HOPS has recently been extended with a novel approach to automatic
allocation of safety requirements to components of a system architecture.
The proposed process can support and simplify the implementation of the
upcoming automotive safety standard ISO26262. A member of the
corresponding ISO committee has been involved in evaluating this proposal
with a view to influencing the refinement and implementation of the
standard (see corroborating evidence 2).
Through commercialisations, EAST-ADL and consultancies funded by Toyota,
the Flemish Mechatronic Institute and others, HiP-HOPS has been taken up
by automotive companies which include Volvo, Toyota, Volkswagen, Daimler,
Fiat, Siemens, Continental, Ricardo and Mecel. HiP-HOPS is being
experimentally used in the design of new active safety systems by these
companies. Daimler have developed their own implementation of HiP-HOPS. In
the aerospace sector, HiP-HOPS is being used by Honeywell and Embraer for
improved system design. New work (2012-present) of potentially high impact
in the sector is focused on harmonising HiP-HOPS with AADL, the emerging
Architecture Analysis and Design Language which is becoming an industry
standard in the aerospace industry. Embraer (see corroborating evidence 8)
have developed their own implementation in HiP-HOPS. The continual and
growing involvement of these companies with HiP-HOPS shows the benefits
they have acquired in streamlining, rationalising and improving dependable
design. Their involvement in over 50 HiP-HOPS publications within the
period, reported case studies (including in refs [1-3] in section 3 and
the eight submitted REF outputs on HiP-HOPS), as well as personal
communication with the industrial contacts cited in section 4, will
confirm these benefits.
3. Societal impact (2008 — present)
Industrial applications in large scale show that HiP-HOPS speeds up the
safety analysis process by automating part of it and thus enabling
multiple iterations of safety analysis that help to improve the design of
safe systems. The impact on safety improvements is difficult to quantify
at this stage, but users largely agree (see industrial contributions to
publications and sources to corroborate impact 6-9) that the process
significantly improves classical safety analysis and influences positively
the safety of systems which can in turn significantly reduce risks of life
losses as well as property and environment damages caused by system
failures
Sources to corroborate the impact
All references to organisations below are supported by factual
statements.
- Contribution to the specification of the EAST-ADL language (Hull is
central in EU projects where language was defined see http://www.atesst.org
and http://www.maenad.eu, founding
members of the EAST-ADL association, Papadopoulos is among two academics
sitting on the 5-member association board, and HiP-HOPS is one of
EAST-ADL supporting tools see association website http://www.east-adl.info/)
- Influence of HiP-HOPS is the new AUTOMOTIVE safety standard, Member of
ISO Committee
- Commercialisation of HiP-HOPS with ITI GmbH (Commercialisation
Agreement available on request from University of Hull).
- Commercialisation of HiP-HOPS with ALL4TEC (Sublicensing and
distribution agreement available on request from University of Hull).
- Data to corroborate the value of sales of HiP-HOPS (Can be provided by
University of Hull).
- Impact on Japanese industry — experience with large clients of
HiP-HOPS, testimony by Managing Director of commercialisation partner
(ITI GmbH), Germany.
- Impact on automotive sector (BOSCH), Functional Safety Manager, Bosch
Automotive (Tier 1 supplier), Germany.
- Impact on aerospace sector (EMBRAER), Manager responsible for
Modelling and Simulation in Embraer and project Leader of in-house
developed HiP-HOPS tool.
- Impact on shipping sector (GERMANISCHER LLOYD), Head of Department
Strategic Research, Germanischer Lloyd.