top of page

HiP-HOPS: Hierarchically Performed Hazard Origin and Propagation Studies

Safety analysis generated from the system design

​

Fault trees and Failure Modes and Effects Analyses (FMEAs) are the backbone of safety assessment in automotive, aerospace, marine and industrial engineering. They are normally produced by hand, in documents kept apart from the design they describe. Each design change can leave them wrong, and on a large system the effort of redoing them decides how often they are redone.

HiP-HOPS removes that gap. Engineers describe how each component can fail, next to the component in the system model. HiP-HOPS then works out how those failures propagate through the architecture and generates the fault trees and FMEA tables for the whole system, again and automatically whenever the design changes.

Because the analysis is automatic, it can be used early and often: while requirements are being turned into an architecture, to decide how strict each component’s safety requirements must be and to compare design options, and later, to show that the finished design is dependable.

HiP-HOPS was conceived by Prof. Yiannis Papadopoulos and has been developed at the University of Hull since 2001.

From architecture model to fault trees

HiP-HOPS works on the architecture models engineers already build, for example in MATLAB Simulink or SimulationX.

Each component is annotated with its local failure behaviour: which failures can start inside it, and how deviations at its inputs, such as a missing or wrong value, turn into deviations at its outputs. The system’s hazards are defined the same way, as deviations at the system’s outputs.

From these annotations HiP-HOPS traces every hazard back through the architecture to its root causes. It produces a fault tree for each hazard, and an FMEA showing, for every component failure, which system effects it can cause.

The HiP-HOPS website walks through a complete example: a brake-by-wire hybrid braking system for an electric vehicle, from Simulink model to fault trees, FMEA and allocated safety integrity levels.

HiP-HOPS analysing a MATLAB Simulink system model, with the HiP-HOPS dialog open over the architecture diagram

​​

Beyond fault trees and FMEAs, HiP-HOPS has grown into a set of connected capabilities. Safety requirement allocation. Working top-down from system hazards, HiP-HOPS decides how strict each component’s safety requirements need to be, automating much of the ASIL allocation and decomposition that ISO 26262 asks for in road vehicles. Pandora temporal logic. Some failures matter only in a particular order: a backup that fails before the main unit is not the same as one that fails after it. Pandora lets fault tree analysis express and evaluate those sequences. Design optimisation. Genetic algorithms search large numbers of design options, such as which components to use and where and how much redundancy to add, for the best trade-off between dependability and cost. Reusable failure patterns. Common failure behaviour can be written once as a pattern and reused, instead of being re-entered for every component.

Architecture model → failure annotations → fault trees and FMEAs → safety requirement allocation → optimised design

HiP-HOPS in industry and standards
​

HiP-HOPS is one of the most established model-based safety analysis methods. Organisations that have taken up the method or the tool include Volvo, Toyota, Honda, Fiat, Continental, Embraer and Honeywell. The tool has been commercialised with partners ITI GmbH, which built it into SimulationX, and ALL4TEC.

The work has shaped standards as well as tools. Through the European projects ATESST, ATESST2 and MAENAD, HiP-HOPS helped define how EAST-ADL, the architecture description language for automotive control systems, describes errors and failures, and the University of Hull is a founding member of the EAST-ADL Association.

The research behind HiP-HOPS now spans over a hundred papers, many written with industrial partners.

Using HiP-HOPS

A commercial version supports fault tree and FMEA generation and analysis, design optimisation and early support for safety requirement allocation. Research licences, including help connecting HiP-HOPS to your own modelling tools, are available directly from the University of Hull: contact Prof. Yiannis Papadopoulos.

Downloads, release notes, the manual and the worked braking-system example are on the HiP-HOPS website.

Developed by Prof. Yiannis Papadopoulos (lead), Martin Walker (Pandora) and Dr David Parker (design optimisation), with colleagues and partners in DEIS.

Talk to us about HiP-HOPS →

← All technologies

bottom of page