Reboot IoT Factory, Phase IPoC PortfolioReboot IoT Factory consortium

16.1.2020

ForewordProof-of-Concepts (PoC) are concrete digitalization experiments carried out in real-world factorycontexts by Reboot stakeholders. These experiments give insight into how factories should moveforward with common digitalization themes called Grand Challenges.

In this PoC portfolio, each PoC experiment from Reboot Phase I is presented as a one-slider.Material is organized based on Grand Challenges. Each one-slider lists business need, solutionapproach and expected business impact. The portfolio is a tool to gain a quick overview of projectactivities within each Grand Challenge.

PoC experimentation work continues in Reboot Phase II, resulting in an additional PoC portfolio.

20 Proof-Of-Concepts and their benefits

Robotics processautomation in production

Digital twintechnical development

Tester predictivemaintenance

Supply chaintransparency and agility

Prospect to project

Wearables inindustry

Automatic componentquality control

Standard robotinterface

Automatic errorhandling

AVR for undestandableinstructions

AI foreman: productivityfrom data

External B2B platform B2B extranet datatransfer

DATA

DRIVE

N SU

PPLY

CHAIN

AND

PROD

UCTIO

NRO

BOTIC

S FUS

ION

LABO

UR AT

DIGIT

ALWO

RK EN

VIRON

MENT

Mobile robot inmaterial handling

Well-being at workAI foreman: Requirementsand core functionality Gamification

Legislation, GDPR, privacyin automatic decisioning

Value of service enabledby digital solutions

Factory acceptance testing

Improved supply response time Lower maintenance costs Less time used in filling IT-systems Customer get €-value for sensored product

Shifting humans to higher value work Double the inspection speed One robot serves multiple production lines Robots becoming more productive

Improved sales predictions

Quick and documented final testing

Supervisors can focus on production instead of resourcing Connected employees work efficiently Wellness improves employee productivity AI according to GDPR

Data-Driven Supply Chain andProduction Management

Data-Driven Supply Chain andProduction Management

Grand Challenge 1Grand Challenge 1

Prospect to Project (P2P)NEED§ Reduce manual non-value-adding work from order-delivery process data flow.

§ Create transparency to prospect data towards all factory stakeholders.

§ Better sales prediction to improve production planning, procurement andbudgeting.

SOLUTION APPROACH§ Design and implement a cloud-based approach for availability of sales

prospect and short-term forecast data for production planning andprocurement.

§ Study the feasibility and create implementation roadmap for making longterm sale forecasts based on cloud-based prospect data.

IMPACT§ Transparency of active sales prospects to the entire organization.

§ Reduced waste and manual work (estimated 84 000 €/yearpotential for re-allocation to value-adding work).

PoC 1.1

TEAM

Technicalrequirements

Productmodules

CRM

Project requirements

Prospectdata

Moduledatabase

Transparency

Factorystakeholders

Supply chain transparency and agilityNEED§ Improve response times of supply chain in terms of production

planning and material flow.

SOLUTION APPROACH§ Design and simulate an optimized approach for material,

information and financial flows within factory’s supply chain.

§ Identify relevant supply chain development points based onprocess mapping and simulated results.

IMPACT§ Development areas in supply chain operations identified and

roadmapped for future piloting.

PoC 1.2

TEAM

CRM

Material

Information

Supplier

Factory(focal firm)

Extranet developmentNEED§ Reduce manual non-value-adding work and increase information

transparency in procurement processes at both factory and its supplychain.

SOLUTION APPROACH§ Define requirements for an integrated supply chain management

platform spanning the factory and its supply chain companies.

§ Scout and evaluate available SME offering in platform solutions andcompare with in-house development activity.

IMPACT§ Increased understanding of internal and supply chain requirements

in terms of extranet functionality.

§ Increased understanding of available solutions and their features.

PoC 1.3

TEAM

Factory(focal firm) Supplier

Researcher

SME

First SCMimplementation

Current extranetsystem

Digital Twin technical DevelopmentNEED§ Improved sensing on thruster unit is needed as enabler for

digital twin, but physical sensor placement in thruster chassisis challenging.

SOLUTION APPROACH§ Study and design virtual sensors as an approach to perform

sensing from challenging deployments. Measurements aretaken at one location and used to estimate values in severalother locations.

§ Study optimizations that reduce the effects of unknowndisturbances to virtual sensors.

IMPACT§ Enable further design and construction of thruster digital twin

in terms of technology and business logic.

PoC 1.4

TEAM

Physical Asset

Digital Twin

Sensordata

On-boardPC

Component load analysis, simulationsand maintenance predictions

Virtual sensors and virtualcomponents

Satelliteconnection

Dashboard

Tester predictive maintenanceNEED§ Tester failures in product testing cause significant delays and

manual maintenance work. Tester performance needs to bemonitored in real-time, and maintenance scheduled predictively andindividually to each tester.

SOLUTION APPROACH§ Study approaches to collect real-time data from tester equipment.

§ Design and implement tester yield analysis based on sliding windowaveraging.

§ Study different HCI methods for delivering real-time notifications totester maintenance personnel.

IMPACT§ Annual labour time savings of around 50 000 € per factory.

§ Improved management of failures (visualization of failures, pareto,top failures etc).

§ Production efficiency and yield improvement.

PoC 1.5

TEAM

TesterWatch

Tester A

Tester B

OK

Repair tester withminimal damages tooutput

No actionsrequired

Robotic process automationin productionNEED§ Product yield management is largely based on manually run daily

reports.

§ Legacy reporting systems poorly support automating of reportgeneration.

SOLUTION APPROACH§ Study the impact of robotic process automation (RPA) for replacing

manual work in production reporting.

§ Scout SME offering for RPA solutions.

§ Pilot RPA approaches with a selected SME (BotLabs).

IMPACT§ Increased understanding of RPA suitability for production automation

and in general factory digitalization cases.

§ Currently saves around 1.5 man-months of work time per year, whichcan be allocated to other productivity tasks.

PoC 1.6

TEAM

Value of Service enabled bydigital solutionNEED§ Digital twin technology is very promising for several digitalization

cases, but its business feasibility especially in terms of product life-cycle management is uncertain.

SOLUTION APPROACH§ Identify a set of key parameters that affect the operational business

impact of a product digital twin.

§ Design and develop a tool for simulating service-based business logicof a product digital twin, based on given starting parameterization.

IMPACT§ Improved means to communicate the customer value of product

digital twin to the potential customer.

§ Increasing understanding regarding the business and earning logicmodels of the digital service solutions.

PoC 1.8

TEAM

Factory Acceptance Test (FAT)NEED§ Reduce manual work in FAT process in terms of parameter

measurement.

§ Reduce need of classification society presence in FAT process.

SOLUTION APPROACH§ Identify points in measurement, where data collection will be

automated.

§ Implement automated data collection in selected points.

§ Study possibilities for cloud-based digital distribution ofmeasurement values for classification society.

IMPACT§ Increased understanding of FAT process automation.

PoC 1.9

TEAM

Grand Challenge 2Grand Challenge 2

Robotics FusionRobotics Fusion

Mobile Robot In Material HandlingNEED§ Reduce manual non-value adding work and increase

productivity in intralogistics.

SOLUTION APPROACH§ Create a concept of mobile robot system and carrying rack that

automates the intralogistics.

§ Factory and chosen SME designed, tested and implementedsystem based on concept.

§ VTT designed and implemented the control system.

IMPACT§ Increased productivity in production: Estimated cost savings are

per factory 45 000 € / year. If the systems lifetime is 5 yearscumulated total cost saving are 225 000 € per factory. Paybacktime will be 2 years.

PoC 2.1

TEAM

MIR

Lift PLCGE

Supermarket

PLCBarcodereader

Production lines

PLC

Control PC

Barcode handlingTask controlAdjustable tasklist UI

GE GE

VTT

Posicraft

Automatic Component Quality ControlNEED§ Increase productivity on robotized assembly cells.§ Reduce manual work related to inspection.§ Increase the quality in inspections.§ Generate digital measurement data from inspections.

SOLUTION APPROACH§ Survey and study feasibility of machine vision methods for detecting

micro-liter droplets of water.§ VTT designed and implemented a prototype based on NIR spectrometry.§ VTT designed the lighting system for NIR-S based inspection.§ Chosen SME (ProtoRhino) provides an integration of the solution to the

production environment.

IMPACT§ Estimated cost savings are 50 000 € / factory / year. If the systems

lifetime is 10 years cumulated total cost saving are 500 000 € / factory.Estimated payback time is few months.

§ Increased accuracy in quality control, increased capacity.

PoC 2.2

TEAM

Standard Robot InterfaceNEED§ Increase utilization rate of cobots.

§ Enable automatic docking / undocking of cobots from individual workstations.

§ Increase the automation rate of assembly.

SOLUTION APPROACH§ Create a concept for a standard interface system between cobot (on manually

movable platform) and a workstation.

§ Create a concept for cobot gripper changer based on workstation needs.

§ Factory and chosen SME implemented, tested and adopted the solution toproduction environment.

IMPACT§ Estimated cost savings are 100 000 € / factory / year. If the systems

lifetime is 5 years cumulated total cost saving are 500 000 € / factory.Payback time will be 1.2 years.

§ New business opportunities for SME (RoBoCo).

§ Increased flexibility of robotized assembly.

PoC 2.3

TEAM

Automatic Error HandlingNEED§ Increase the productivity on robotized assembly cells.

§ Reduce downtime of robots that have ended in an error state.

§ Have robots automatically recover from error states.

SOLUTION APPROACH§ Study video analytics as a way to identify and analyze

categories of error situations.

§ Tentative testing conducted.

IMPACT§ Increased understanding of video analytics feasibility on error

recognition, to be used in future solution planning anddevelopment.

PoC 2.4

TEAM

Labor at DigitalWork Environment

Labor at DigitalWork Environment

Grand Challenge 3Grand Challenge 3

Requirement specifications & corefunctionality for AI foremanNEED§ Produce automated production planning proposals based on

multiple data sources to support faster day-to-day decision-makingregarding production planning.

SOLUTION APPROACH§ Specify requirements and functionality for AI-powered generation of

production planning proposals, titled ‘AI foreman’.

§ Study how current manufacturing execution systems (MES) shouldbe improved to enable cloud-based AI production planning.

§ Scout SME offering for the proposed need.

IMPACT§ ABB Adafo MES system modified for data extraction and insertion.

§ Employee competence matrix maintenance automated in MES,estimated 20k€ annual savings.

PoC 3.1

TEAM

AdafoMES

VisualFrontend

Events Example: Worker X completed event Y

Worker data Example: Worker X received 10 points for completion of event Y

CompetenceMatrix

Virtual JobSupervisor

VR to achieve moreunderstandable instructionsNEED§ Lessen the need for experienced assembly workers to spend time

instructing new employees on manual assembly tasks.

SOLUTION APPROACH§ Utilize game engine technology to iteratively develop and evaluate

training applications in virtual reality (VR).

IMPACT§ Increased understanding on how to use VR technology to train new

assembly tasks to employees.

§ Lessen the need for experienced employees to participate in trainingsupervision (~3 man months per year).

PoC 3.2

TEAM

AI Foreman: Productivity from data

NEED§ Production resourcing is a complex optimization task that is difficult

to solve manually in changing conditions.

§ Foremen need tools that can propose resourcing automaticallybased on multiple data sources.

SOLUTION APPROACH§ Study different optimization algorithms.

§ Study the optimal combinations of data (production data, orders,storage status, personnel availability, line capacity, competencematrices and wellbeing) to be used to automatically schedule dailytasks and capacity planning.

IMPACT§ Potential to save a lot of working hours: both working hours of

human foremen and employees.

PoC 3.3

TEAM

Wearables in industryNEED§ Efficient methods to inform employees about forthcoming tasks in

the future digital work environment.

§ Assess the potential of smartwatches in measuring well-being atwork.

SOLUTION APPROACH§ Smartwatch notifications can be used to inform employees what

they need to do during the next 5 minutes.

§ Study how to use smartwatch sensors to measure employees stressthrough applying machine learning algorithms.

IMPACT§ Faster reaction to problems -> less breaks at the automated

manufacturing lines.

§ Enable production planning in terms of employee well-being.

§ Enable AI powered production planning.

PoC 3.4

TEAM

Wellbeing at workNEED§ Studies show that productivity of factory goes hand in hand with

employee wellness. Therefore, it is important to study how wellbeingat work can be measured and improved.

SOLUTION APPROACH§ Measuring and improving employees mental and physical load

through questionnaires.

§ Install HappyOrNot machines on the factory floor and instructingemployees to use it.

§ Collecting and analyzing HappyOrNot data to identify correlations toproduction and other data.

IMPACT§ Increased understanding on what affects well-being and how it could

be improved.

§ Enable construction of employee digital twin based on well-beingdata.

PoC 3.5

TEAM

Gamification for Factory floorNEED§ Improving work motivation and employee satisfaction.

§ Increasing attractivity of the factory work for the next generation ofemployees.

SOLUTION APPROACH§ Design guidelines for gamification for factory floor.

§ Virtual currency tied to work tasks.

§ Gamified assembly instructions.

IMPACT§ Increased motivation of employees.

§ Increased involvement of employees in change.

PoC 3.6

TEAM

LegislationNEED§ Understand what conditions data privacy legislation (such as the

GDPR & the Act on the Protection of Privacy in Working Life) sets forpersonal data gathering and processing.

SOLUTION APPROACH§ Systematic review and analysis of the relevant EU and national-level

legislation, with respect to the current data processing operations,future data needs and possibilities to collect data.

§ Creation of AI Foreman Regulation Framework for support andguidance for implementation of legislative requirements in AIForeman.

IMPACT§ Lowered threshold for factories to implement employee data

collection.

§ Increased understanding of legislation sets for personnel dataprocessing.

PoC 3.7

TEAM