October 2016

Presentation of the Mantis Project at Sirris’ seminar: fleet-based analytics for data-driven operation and maintenance optimization

On October 24th Sirris organized an industrial seminar on the opportunities and challenges related to fleet-based data exploration. During this seminar, a general introduction to the MANTIS project was first given, followed by presentations from several partners within theMANTIS project including: the Mondragon University (press machines), the Eindhoven University of Technology (shaver manufacturing), 3E (Photovoltaic Plants), Ilias Solutions (Vehicles), Atlas Copco (compressors) and Sirris. The event was a real success with around 45 participants and offered participants via real-world use cases in the different industrial domains mentioned above the opportunity to see how data-driven analytics on a fleet of machines can optimize the operation and maintenance of those.

Sirris presentation MANTIS
Tom Tourwe introduced the MANTIS project

 

Mondragon presentation MANTIS
Urko Zurutuza presented the Press Machine Maintenance Use-Case

If you would like to have further information on the outcomes of this seminar, please contact Caroline Mair (caroline.mair@sirris.be)

Deep learning for predictive maintenance

There are two extreme approaches to predicting failures for predictive maintenance. The white box approach relies on manually constructed physical and mechanical models for predicting the failures. The black box approach, on the other hand, relies on failure prediction models constructed using statistical and machine learning methods based on the data gathered from a running system. The figure below illustrates such data driven failure prediction for a machine monitored by three sensors.

data driven failure prediction
data driven failure prediction

Machine learning algorithms are used to identify failure patterns in the sensor data that precede a machine failure. When such patterns are observed in operation, an alarm can be triggered to take corrective action to prevent or mitigate the eminent failure. For example, failure predictions can be used to optimize the maintenance actions, such as scheduling the service engineers or managing the spare parts storage to reduce the downtime cost.

Automatic feature extraction

An important part of modeling a failure predictor is selecting or constructing the right features, i.e. selecting existing features from the data set, or constructing derivative features, which are most suitable for solving the learning task.

Traditionally, the features are selected manually, relying on the experience of process engineers who understand the physical and mechanical processes in the analyzed system. Unfortunately, manual feature selection suffers from different kinds of bias and is very labor intensive. Moreover, the selected features are specific to a particular learning task, and cannot be easily reused in a different task (e.g. the features which are effective for predicting failures in one production line will not necessarily be effective in a different line).

Deep learning techniques investigated in the MANTIS project offer an alternative to manual feature selection.  It refers to a branch of machine learning based on algorithms which automatically extract abstract features from the raw data that are most suitable for solving a particular learning task. Predictive maintenance can benefit from such automatic feature extraction to reduce effort, cost and delay that are associated with extracting good features.

Sirris seminar on fleet-based analytics for data-driven operation and maintenance optimization

On October 24th Sirris is organizing in Belgium an industrial seminar on the opportunities and challenges related to fleet-based data exploration. During this event Belgian as well as other European industrial partners from the MANTIS project will present their experience with fleet-based analytics (based on use-cases from the MANTIS project).

Many companies operate a fleet of machines that have a similar, almost identical behaviour in terms of internal operation, application and usage, such as for example windmills, compressors and professional vehicles. This set of almost identical machines is defined as ‘a fleet’.

In addition, more and more, those machines are equipped with several (smart) sensors, that can capture data on operational temperature, vibrations, pressure and many other features, depending on the machine. In addition, the communication and data storage technologies are becoming ubiquitous, making it possible to gather the data in a central platform and derive insights into normal and anomalous behaviour across the entire fleet of machines. By comparing for example the behaviour of a single machine to the rest of the fleet, one can identify if a machine is underperforming due to misconfiguration or imminent failure. The analysis of this data can also help service and maintenance personnel to have a more detailed and optimised maintenance planning, e.g. ensuring an optimal distribution of the entire fleet in terms of remaining useful life, in order to manage the work load of the service engineers. Therefore, the exploitation of the data collected on a fleet of machines is a real asset for maintenance and service personnel and, at a larger scale, for an entire company.

You are interested in this event? Check out the event’s agenda and register here

Programme

13:00 – 13:15: Registration and coffee

13:15 – 13:30: Setting the scene (Sirris)

13:30 – 14:30: MANTIS project: Cyber Physical System based Proactive Collaborative Maintenance

    • Project goals and challenges by Sirris
    • Fagor use case (press machines) – title to be announced
    • Philips use case (shaver manufacturing) – title to be announced

14:30 – 15:30: Root cause analysis

    • Barco – Vitriol: let open source data science talk quality and business at Barco Projection
    • 3E – Data-driven Fault Detection for Photovoltaic Plants: Data Quality, Common Faults and Data Annotation
    • Atlas Copco – SMARTLINK & root-cause analysis on compressors worldwide to improve on operational efficiency

15:30 – 15:50: Coffee break

15:50 – 17:10: Failure prediction & Operational optimisation

    • Barco – LightLease Predicting Lamp Behaviour in Digital Cinema
    • Ilias – Towards Predictive Vehicle Fleet Management
    • Maintenance Partners – Performance optimisation and failure prediction of wind turbines
    • Pepite – Analytics for operational optimisation

17:10 – 17:30: Closing remarks (Sirris)

17:30 – Networking reception

MANTIS-Platform Requirements or How to Make Sure the Platform Fulfils Partner’s Needs

One of the most important tasks to ensure flawless working on different work packages is to have fully consolidated requirements towards the MANTIS platform. Not only is it important to have all different requirements aggregated, but also to know in which project tasks those requirements will be addressed, by whom and how to manage those during the project lifetime. To handle this challenge with excellence, MANTIS partners decided to do a 4-step approach in requirement definition towards the MANTIS platform.

Defining User Scenarios & Deriving Requirements

In the MANTIS project, we do have specific use cases we later on want to test the platform against. Each of these use cases detailed their scenarios and related them to the MANTIS objectives. From there on, requirements in the form of a table were compiled. Those requirements cover functional, non-functional, technological and business needs. As a result, we had the first form of requirements which – of course- still needed unification and consolidation.

Extending Maintenance Use Case Requirements by MANTIS Partner Requirements

This task focused on what technology can and should be used for the user scenarios. The main task members therefore were the technology providing partners. We also expected requirements originating from technology push of the providers. Those additional requirements were the second form of requirements we had which – again, of course – still needed unification and consolidation. Nonetheless, correlations to the user scenario requirements were already identified and noted.

Requirements and Project Objects Consolidation

The platform requirements and the user scenario requirements were revisited for consistency and updates. Furthermore, a refinement of the requirements was performed based on the early sketches of the MANTIS architecture. As a result, 45 different requirement categories were identified, 27 of those categories were then identified as “not to be addressed inside of MANTIS”, since those categories mainly were basic requirements towards a general platform architecture and not MANTIS specific.

 

MANTIS requirements categories
MANTIS requirements categories

 

Refinement of the Requirements

As the last task, which is currently still running strong, we wanted to match all those 900+ requirements that were identified and defined to the 45 defined requirement categories. At the time of writing this article, the matching is completed and 900 requirements could be reduced to about 330 MANTIS specific requirements. Also, the MANTIS specific requirement categories were matched to the different project task, to make sure that the requirements will definitely be addressed.

So what’s ahead: Definition of MANTIS platform requirements is near its end. What’s left to do is to choose a good way to manage the requirements during projects lifetime. As it seems, this will be done by separating Excel sheets according to tasks and categories. This will keep management of requirements handy and easy.

 

Matching Requirements to MANTIS requirement categories
Matching Requirements to MANTIS requirement categories

An integrated Mantis-Platform for Off-Road and Special Purpose Vehicles

Introduction

Off-Road and Special Purpose vehicles are used all over the world in various environmental conditions. They exist in all different kinds and formats and, within companies, a broad range of types of such vehicles will be in use.

Maintenance on these vehicles, be it preventive or corrective, can cause unavailability, having a negative impact on both productivity and efficiency. An overall objective regarding maintenance is to maximize the availability of the vehicles at the lowest maintenance cost. Therefore, a pro-active and preventive maintenance approach should lead to important savings, with higher availability.

The ILIAS approach

Most of these vehicles are already equipped with on-board HUMS systems/black boxes. The data generated by these on-board systems contain a broad range of information that can be used as input in a MANTIS-based platform in order to optimize the full maintenance strategy.

The diversity of HUMS systems, however, is very broad, even on the same type of vehicles. Each vehicle has its own configurations, interfaces, data formats, etc. Hence, there is a need to convert the collected data from various systems into a uniform and structured format in order to make them further exploitable.

This observation has led us to the conclusion that there are two viable approaches to building MANTIS-based platforms:

  • A per-vehicle type / HUMS system platform approach, aiming at an optimum maintenance strategy for a small number of equipment types.
  • An open platform approach that can be easily customized by the user to the type of vehicle/HUMS system(s) being used.

We opted for the second approach but have not limited it to the collection of data only but broadened it to a complete set of functionalities within the MANTIS-based platform.

Based on the experience and know-how gained from the collaboration within the MANTIS project and the architectural guidelines derived from it, ILIAS Solutions aims at building a platform that provides a complete solution from the readout of the black box until the optimization of the maintenance plans in an environment with high numbers of highly complex and mobile assets.

The ILIAS platform, therefore, provides users with an integrated set of user-friendly tools, permitting them to:

  • Import data from external sources like ERP systems, leading to a centralized data set. (Step 1)
  • Import raw data coming from any kind of HUMS system and cleanse them, based on automatic data wrangling, leading to state detection and health assessment. (Steps 2, 3, 4)
  • Make analysis of the data via different algorithms and translate them into rules/conditions to apply in the system. (Steps 5)
  • Define rules/conditions, including use and abuse rules, for triggering maintenance or other linked actions, based on the combined dataset. (Step 6)
  • Approve/disapprove the system-proposed maintenance actions and register them to make the system self-learning. (Steps 7, 8, 9)

This figure illustrates the approach.

MANTIS Approach proposed by ILIAS as a maintenance platform in Off-Road and Special Purpose vehivles
MANTIS Approach proposed by ILIAS as a maintenance platform in Off-Road and Special Purpose vehivles

The figure below illustrates how we go through different steps in implementing the platform, following more iterations to improve the system.

ILIAS is running different steps in implementing the platform
ILIAS is running different steps in implementing the platform

Conclusion

For Off-Road and Special Purpose vehicles, the overall objective regarding maintenance is to maximize the availability of the vehicles at the lowest maintenance cost. Thus, a proactive and preventive maintenance approach leads to important savings, with higher availability.

ILIAS Solutions aims at building a platform that provides a complete solution from the readout of the black box until the optimization of the maintenance plans in an environment with high numbers of highly complex and mobile assets.

This platform should be an open platform that can be easily customized by the user to the type of vehicle/HUMS system(s) in use and where a number of rule sets/conditions are defined in a user-friendly way.  This allows the system to trigger predictive maintenance actions. Analysis of broad data sets will lead to additional rules and conditions, optimizing the platform it selves.