Smart-Troubleshooting in the Connected Society

Smart-Troubleshooting in the Connected Society

Francesco Flammini, Diana Unander, Welf Löwe, Shiyan Hu, Mauro Caporuscio, Morgan Ericsson, Narges Khakpour, Diego Perez-Palacin, all from the LNU Computer Science Department, and the company Sigma

Technology expressed their interest on an activity addressing troubleshooting of smart-devices in the modern highly interconnected society. The total budget is 200.000 SEK, of which 100.000 SEK are self-funded by the researchers own time committed to the project, therefore the SEED co-funding would be 100.000 SEK. The expected duration is 4 (fours) months. The aim of the activity is to conduct a preliminary study addressing the state-of-the-art and a possible framework architecture to be used to develop a project proposal for research funding applications.

Background, Industrial Value, Goal

Sigma Technology is part of Sigma and is a global supplier of product information, software and embedded solutions, and offshore development. They have more than 500 employees in their offices located in Sweden, China, Hungary, the USA, and Ukraine. Sigma is owned by Danir AB that has about 3,600 employees and a network of partners in twelve countries.

Sigma Technology has a strong interest in developing research in order to solve the issues related to troubleshooting in case any malfunctions occur in systems including several connected devices. Connected devices generate error logs, and when several devices from different vendors fail to communicate, troubleshooting becomes a challenge. This research project aims to find a solution to this by gathering error logs, analysing them, and matching them with both software fixes and proper troubleshooting instructions.

The project is connected to research areas and domains within Smart-Cities, Internet of Things and People, Automatic Transportation Systems (ATS), among others.

Therefore, the project is expected to provide a high industrial impact and societal value considering the relevant and increasingly critical applications addressed by the research project, like ATS. Furthermore, it addresses many of the current scientific challenges including big data analytics, machine learning, cloud computing, sematic modelling, and cyber-security, which fall within the interests and expertise of the researchers involved in the proposal.

As a side effect of the proposal, three of the researchers involved are new to the department, so working together will be important for them to learn about each other’s strengths and competences to be used in the project. Overall, the group of people involved include 3 senior lecturers, 2 docents and 2 full professors.

Finally, there is a strong connection between this project and the SYNERGY project that has already been funded within the KK-stiftelsen Synergy Call-for-Proposals, especially regarding the self-adaptive aspects enabling dynamic reconfiguration and thus self-healing, and the decision-support approaches needed to reason under uncertainty.

Research Question, Objectives, and Expected Results The seed activity poses the following objectives:

O 1. How should be the functional architecture of a system able to collect and analyse device error log files and product information, possibly trans-code them in machine-readable formats, to be used in “troubleshooting cloud” for subsequent matching to troubleshooting solutions?

O 2. Which are the research areas and specific activities to be developed in order to achieve the objectives of the research project? How are they interconnected? Which applications can be used to validate the approach?

Activities

In order to achieve the aforementioned objectives, the following activities will be performed:

A 1. Critical review of the state-of-the-art and development of a detailed reference architecture including all the modules that are needed to mine, store, retrieve, interpret and apply all the relevant troubleshooting information based on heterogeneous sources and tackling the issues related to incomplete and unstructured data. – O 1 A 2. Definition, detailed description and classification/grouping of the research activities addressing semantic

modeling, domain specific languages, model-driven engineering, formal-methods, machine-learning, big-data

analytics, etc., that are considered valuable for the development of the research project, assigning precise roles,

scopes and responsibilities. Exploration of grant application opportunities considering both national research

funding (e.g. KK-stiftelsen) and EU Open Calls (e.g. H2020 ECSEL Joint Undertaking). – O 2

Consortium

Companies and research partners involved in the seed activity are:

1. Sigma Technologies – mainly involved in defining industry requirements and case-studies 2. Linnaeus University, Dept. of Computer Science – definition of research activities and results

Other industrial partners and end-users involved in the research will be defined later during the development of the detailed project proposal. Current suggestions of potential partners include: Södra, Electrolux, and Toyota Material Handling. Researchers from other fields or with other competences might be defined as well.

Time Plan

This seed project will last for 4 months starting in March 2018. The activities A1 and A2 are mainly sequential.

Month

Activities 1 2 3 4

A1 A2

Budget

The overall budget of the project is approximately 200.000 SEK. We are applying for 100.000 SEK co-funding since the remaining 100,000 SEK will be covered by the department of Computer Science at LNU (researchers own time).

Main costs are reported in the following table.

Description Amount Details

Actual costs 40.000 SEK Expenses and materials needed for project development, travels and workshop organization, coordination, evaluation, reporting, action planning, research proposal management, etc.

Project Assistant

60.000 SEK Money to be paid to a Project Assistant to support the preliminary development of the research project (e.g. by a systematic review of the state-of-the-art).

Total 100.000 SEK

Specific efforts and contributions (tentative) of people committed to work on the project:

 Francesco Flammini: Technical Manager, Leader – Technology Integration

 Diana Unander: Project Coordinator

 Welf Löwe: Project Supervisor

 Shiyan Hu: Project Reviewer, Leader – Industrial Applications

 Mauro Caporuscio: Leader – Context-Awareness and Self-Healing

 Morgan Ericsson: Leader – Log Analysis and Machine-Learning

 Narges Khakpour: Leader – Formal Modelling and Model-Checking

 Diego Perez-Palacin: Leader – Domain Specific Languages

 SIGMA: Leader – Product Information

1 The budget allocated to the different entries could be possibly be re-distributed, e.g. if the student grant could not be assigned.