Available online at www.sciencedirect.

com

ScienceDirect
Procedia Manufacturing 11 (2017) 372 – 379

27th International Conference on Flexible Automation and Intelligent Manufacturing, FAIM2017,

27-30 June 2017, Modena, Italy

Criteria definition for the identification of HRC use cases in

automotive manufacturing
Alessandro Zanellaa*, Alessandro Cisia, Marco Costantinoa, Massimo Di Pardoa, Giorgio
Pasquettaza and Giulio Vivoa
a
Centro Ricerche FIAT SCpA, Strada Torino 50, Orbassano 10043, Italy

Abstract

The study aims at the definition of a methodology for the objective identification of the most suitable applicative use cases for a
profitable exploitation of HRC technology. The analysis is based on the preliminary assignment of values to multiple Key
Parameters (KPs). The KPs identification is based on a methodological analysis applied to multiple manufacturing cells in
production. Core of the process was the identification of the criteria and the KPs. A systematic application of the tool was made
to test and fine-tune the developed methodology. Criteria and methodology that were defined in the study are summarized.

©©2017
2017The
TheAuthors. Published
Authors. by by
Published Elsevier B.V.B.V.
Elsevier This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the 27th International Conference on Flexible Automation and
Peer-review under responsibility of the scientific committee of the 27th International Conference on Flexible Automation and
IntelligentManufacturing
Intelligent Manufacturing.

Keywords: Human Robot Collaboration; industrial application; methodology; automotive; use case identification.

* Corresponding author. Tel.: +39-011-90-83336; fax: +39-011-90-83666.

E-mail address: alessandro.zanella@crf.it

2351-9789 © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the 27th International Conference on Flexible Automation and Intelligent Manufacturing
doi:10.1016/j.promfg.2017.07.120
 Alessandro Zanella et al. / Procedia Manufacturing 11 (2017) 372 – 379 373

1. Introduction

Human Robot Collaboration is a rapidly emerging technology which is expected to have an important impact on
future manufacturing design approach in heavy industries such as the automotive manufacturing. Many reasons
concur to create this expectation. First of all there is the possibility to exploit the physical abilities of the robot such
as precision, repeatability and force together with the human operator cognitive (intelligence, problem solving,
improvisation, immediate vision) and physical (manipulation, dexterity) capabilities. Secondarily it is possible to:

• reduce the ergonomic issues using the robots to carry over heavy operations;
• improve quality by:
o robot’s characteristic repeatability;
o introduction of controlled adaptive constraints in the operator’s activity;
• improve productivity using the robot to perform “Non Value-Added Activities” (NVAA) instead of the operator;
• give support to elderly or reduced work capacity operators and reintroduce them in the workforce, and so on.

While the expectation is thus motivated, both the technology and the regulatory framework matured just recently.
The technical regulation was defined in 2011 in the standard EN ISO 10218-2:2011 (Robots and robotic devices.
Safety requirements for industrial robots. Part 2: Robot systems and integration) [1], that set the allowed behavior in
Human Robot Collaboration applications. Nevertheless the standard itself quoted that “Additional information and
guidance on collaborative robot operations will be contained in ISO/TS 15066 (currently under preparation)”. The
ISO/TS 15066 [2] (Robots and robotic devices - Collaborative robots) has been published just recently in 2016 and it
is setting the limits and methodologies for the safety in the workplace in Human Robot Collaboration (HRC)
applications. In parallel to EN ISO 10218-2:2011, the EN ISO 10218-1:2011 [3] was published to set the hardware
and functional characteristics that a collaborative robot has to fulfil. After the publication of this robot’s standard the
commercialization of collaborative robots began starting by low duty robots (Universal Robots UR5 and UR10 [4]
respectively with a payload of 5 and 10 kilograms; and KUKA LBR iiwa [5] with a payload of 14 kg), and going up
to higher payload collaborative robots such as the FANUC CR-35iA [6] (35 kg) and the COMAU AURA [7] (110
kg) introduced at the AUTOMATICA 2016 fair.
The current availability of robotic technologies and the completeness of the regulatory framework allow the
implementation of effective HRC application in production. Nevertheless, while the ISO standards are required for
the proper design of the workplace and the cell, the design and use of a HRC application in production has to be
motivated by a proper benefit analysis. In facts, while currently many use cases are declared and tested, their
identification process is often just based on an experience analysis. The different approaches to the workspace and
the design of the application itself are inherently modified by the collaboration between human and robot that, up to
now, was not allowed to be used in the industrial applications.
The identification and choice of workplaces and applications that can take proper advantage of the HRC
technology is not trivial and requires the comparison with the currently used technology which is based to traditional
approach where the operator and the robot cannot share the same space and cannot co-operate on the same
application.
Other studies cope with the definition of a quantifiable methodology for the identification of suitable
collaborative cells; for instance Teiwes, et Al. [8] propose an adjustment of the MTM analysis in which an added
overall score is calculated through the assignment of a specific score to each kind of operation defined in the MTM
analysis. This work gives an interesting approach to a modified MTM, but doesn’t cope with the definition of a
proper set of KPIs. On the other hand Grahn, et Al. [9] suggest a set of KPI for the evaluation of potential benefits
provided by a cooperative assembly workplace.
The performed study aims at the definition of a methodology for the objective identification of the most suitable
applicative use cases for a profitable exploitation of HRC technology. The analysis is based on the assignment of
values to multiple Key Parameters (KPs). The KPs identification is based on a methodological analysis applied to
multiple manufacturing cells in production. Core of the process was the identification of the criteria and the KPs. A
systematic application of the tool was made to test and fine-tune the developed methodology. The paper wants to
summarize the criteria and methodology that were defined in the study.
374 Alessandro Zanella et al. / Procedia Manufacturing 11 (2017) 372 – 379

2. HRC applications identification process

The necessity to perform an identification process arises from the will to apply the HRC technology on an
application in order to show at its best all benefit that the technology can give, while causing the minimum possible
impact on the production line. The developed methodology has been systematically applied during the analysis of
assembly-shop, but has been applied to the analysis of the body-shop afterwards.

2.1. Requirements and initial assumptions

The methodology is developed with the goal to:

• identify cells/applications for an effective and fruitful deployment of HRC technology;

• make an objective and quantitative evaluation of cells which could take advantage of HRC concepts;
• evaluate and score effective feasibility and draft benefit analysis against a set of criteria collected in a limited
number of KPs;
• be applied to existing applications and cells (not green field case).

During the initial data collection phases, to be performed in the manufacturing line, two main class of know-how
are required, i.e. Know-how on “As Is” plant layout and performances of existing cells and processes, and HRC and
robotic systems know-how; this know-how separation is kept in the method’s definition and data collection phases.

2.2. Main Method Structure

The general Structure of the method is summarized in the following Figure 1. The main structure is divided in
two phases: Phase A which comprises the data collection and a first filtering of the cells according to input
parameters and performance analysis (which means for example that there is no need to apply HRC to a profitable
automatic application or that a “red” ergonomic cell can take higher advantage by the application of HRC), and
Phase B for the evaluation of feasibility and benefit analysis. A specific attention was given in Phase A to the
identification of KPs which are already available in the descriptive tools used in the plant in order not to require
further data acquisition or elaboration effort. The following schematic bullet list is a summary of the main
characteristics of the two Phases.

• Phase A:
o The goal is to reduce the initial set of data to a useful amount and to identify cells which can strongly benefit
from HRC technology;
o It requires only the know-how on the existing process and plant layout;
o Tools, KPs used and adopted criteria are not unique but depend on the target and scope of the analysis (e.g.
identification of best HRC cell in a new plant or best HRC cell in an existing paint shop with the aim to
improve quality); a fixed set of KPs could be counterproductive;
o KPs are identified and evaluated according to how the benefit will be evaluated/validated;
o Requested data have to answer to the following questions:
a) Which cell/application (identification data);
b) How can the application be described;
c) PROs: which characteristics of the application or workspace can be improved by the use of HRC;
d) CONs: which characteristics of the application or workspace can affect negatively the application of HRC;
e) Benefits: which quantifiable benefit the HRC application can give,
o
o KPs need to give a quantified evaluation of parameters in order to answer to the above questions;
o Not all the required KPs are collected for the evaluation in this phase;
o Additional KPs and more and more detailed information are collected recursively in the Phase B;
 Alessandro Zanella et al. / Procedia Manufacturing 11 (2017) 372 – 379 375

o The initial selection can be performed on the base of the internal manufacturing evaluation method (e.g. in
FCA most input can be collected from the World Class Manufacturing pillars which collect and analyze
indicators such as operator’s performances, ergonomics , quality, safety and so on);
o Additional inputs like quality issues (caused by Human errors due to fatigue, distraction…), logistic support
and type information and so on.

• Phase B:
o The goal is to reduce the initial dataset to a minimum amount by the identification of the most suitable
workcell or application for the HRC implementation in terms of ergonomics improvement, HRC exploitation
and application, technological feasibility, impact on logistics by the new application, or impact caused by
logistic to the new application and so on;
o It is an iterated phase with more and more detailed analysis to narrow the input dataset to a desired amount of
applications;
o It requires the concept design of the application to be realized according to the Human Robot Collaboration
robotic technology and technical standard application;
o The final output results furnishes an initial concept design of the candidate applications;
o The used KPs derive from the first phase;
o It doesn’t require any further data collection from the plant and involves mainly the evaluation and assignment
of scores to a set of calculated KPIs;
o KPs and adopted criteria depend on the target and scope of the analysis;
o The calculated KPIs are weighted according to the target and scope of the analysis as well;
o The final analysis’ result is based on a score obtained by the sum of different KPIs.

The methodology is used to give an objective score to a set of candidate applications, nevertheless it is important
to highlight that the obtained score cannot be absolute since the set of KPs used and the weight assigned to their
evaluation in the score assignment is dependent on the target and scope of the analysis. In the rest of the paper the
description of the method applied is focused rather on the Criteria and KPs used for the evaluation than on the
absolute value assigned to a specific KP or to the weights applied to specific set of KPs.

Fig. 1. Main methodology structure

376 Alessandro Zanella et al. / Procedia Manufacturing 11 (2017) 372 – 379

3. Initial data collection and scoring- Phase A

Phase A requires a collection of data from the plant as an input. For each analyzed cell a set of data is collected in
a table which describes the application. The required fields are grouped in the following clusters, each describing a
characteristic aspect of the analyzed application or cell:

• cell identification and description: as reference and for support to operators in following analysis;
• ergonomics and related tools: description of the current ergonomics and tools which are used in production to
improve it. These data are used in order to better rank the applications in which ergonomics for the operators is
currently critical, or achieved by expensive tools and supports. HRC can improve in these cases of the current
situation;
• operator’s position and room availability: description of the position of the operator, for example, in respect to
the vehicle and room availability (for hands and for body) during the operations in order to consider difficulties
and risks arising from the co-working in restricted areas; space in the cell/area in order to evaluate if the robot
can be placed without hindering operative and safety areas for the operators;
• conveyor description: description of the conveyor type and its use (stop station, continuous moving...). These
factors can affect both the technical complexity and the difficulties for the operator in relation to the application;
• operating time: Takt time, duration of the reference operation, NVAA (Non Value Added Activities) percentage;
these data are originated from the MTM-UAS (Methods-Time Measurement - Universal Analysis System)
analysis of the cells;
• type of logistics: information about logistics for the specific cell in order to consider impact on the new cell’s
layout coming from the presence of the robot.

Inputs are collected in a table, see Figure 2 as a partial example, and scored (the lowest the best). A ranking is
obtained and it is used to restrict the initial dataset. After Phase A it is possible to decide which applications can be
analyzed in Phase B and which can be discarded.

Fig. 2. Extract of an input table from Phase A – Example of application description and field allowed values
 Alessandro Zanella et al. / Procedia Manufacturing 11 (2017) 372 – 379 377

4. Phase B: cell concept design and KPs assignment

Phase B requires knowing how the application will work according to HRC criteria. It requires both technical
know-how and know-how on technical standards and regulation. According to Figure 1, the initial steps are:

• A draft hypothesis of the main phases made for each potential application. The description level gets recursively
more and more detailed; each iterations of the main cycle reduce the dataset and increase the details;
• A concept design of the cells defined in terms of required functionalities and thus in terms of systems and
hardware.

The following steps in the analysis require the assignment of the main KPs through the description of the
characteristics of the cell under development. The following main aspects (each one corresponding to a main key
parameter - KP) are considered:

• Technological complexity
• HRC Relevance
• Benefits/Costs indicator
• Ergonomics & Safety
• Logistic Interface

The final overall score is based on the sum of the above listed KPs.

4.1. Technological complexity index

From a technical point of view most of the HRC hypotheses are realizable safely by properly adding safety
systems, sensors, complexity in the logic and so on.
The Technological complexity KP considers the technologies required in terms of:

• Sensors
• Actuation systems
• Software architecture
• Software complexity
• Additional Hardware requirements, and so on

A proper description of the application’s technology will score the KP giving an evaluation of the technological
complexity and feasibility of the application. Besides of the feasibility itself, the technological complexity of the
system will have consequences:

• Reliability
• Costs
• Risk assessment complexity

The evaluation of technological complexity is comprised in the evaluation of “Which characteristics of the
application or workspace can affect negatively the application of HRC?”

4.2. HRC Relevance

HRC Relevance is a KP value assigned to weight the effective use of all the HRC phases and modalities. It is not
required that an application uses all the possible features, but, for example, designing the application for a pure hand
guiding could probably result in the use of a Partner with a potential economic saving, while no concurrent
technology can replace a HRC robot in case the application requires more different phases.
378 Alessandro Zanella et al. / Procedia Manufacturing 11 (2017) 372 – 379

4.3. Benefits/Costs (B/C) indicator

This evaluation is not a complete B/C evaluation, but a preliminary one to support the decision process. The KP
contains budgetary cost information on the realization of the cell and on the achievable savings when the application
is running, and productivity analysis data. Some of the productivity data are coming from a modified MTM-UAS
[10] analysis, defined considering the motion of the robot in parallel with the operator together with its interaction.

Fig. 3. Representation of the modified MTM-UAS analysis. Sectors a), b),and c) are explained in the text

Figure 3 represents a modified MTM analysis. It is mainly composed by three parts: a) representing the analysis
of the operations as performed by the operator alone; b) with the analysis of the operations performed by the human
operator in collaboration with the robot (represented in the sector c). On the left side of the analysis each operation
is classified and quantified, for the operator, according to the MTM analysis (VAA, NVAA and subclasses such as
walk – KA, Transform – TR, passivity – PA, and so on); while for the Robot the classification is in term of
interaction, wait, hand guiding, handling, automatic and so on, and in terms of cooperative phases according to
ISO/TS 15066 [2] (SSM, SMS, HG, PFL) plus the stop condition.
The procedure furnishes many information used in the evaluation of the B/C indicator but its completion is time
demanding and can be fruitfully applied only to a selected set of applications.
 Alessandro Zanella et al. / Procedia Manufacturing 11 (2017) 372 – 379 379

4.4. Ergonomics & Safety

A single score KP collects information on Ergonomics and Safety. The ergonomics evaluation takes into account
both the ergonomic level of the AS-IS reference application and the TO-BE expected ergonomics in the HRC
application. Other parameters related to the safety of the operator both in the AS-IS and in the TO-BE cases are
considered, such as the evaluation of the operator’s workplace in reference to the vehicle’s chassis and considering
generically clearance for the body and the hands in order to reduce the crushing and punching risk.

4.5. Logistic Interface

Finally, a single score is assigned for the logistic impact considerations.

The considered parameters are Side line logistics type, Logistics container position, Logistics pickup type,
Multilayers and separators in container, singularly packed parts, presence of variants in the product (requiring both
more complex grippers and more complex automation of the logistic toward the robot).

5. Final considerations

After the evaluation, the main Key Parameters are summed with their relative weights. The methodology
furnishes a list of selected applications with their relative scores.
As stated, Phase B is cyclic, and it can be executed until the final dataset is restricted enough, i.e. until a single
application is selected or until a sufficiently restricted set of applications is obtained. As from the initial
requirements the methodology should be objective, yet the scoring phases depend on the assignment of values
performed by an expert; furthermore the assigned value require a concept design which can be correct only after the
full design of the cell. The cyclic execution brings to a more detailed concept design and layout at each iteration of
the analysis. This incremental definition of detail is also a support to the final design of the application.

6. Conclusion

A methodology was created in support to the identification of suitable industrial applications in the automotive
field for a fruitful application of the Human Robot Collaboration approach. The methodology requires the cyclic
investigation and analysis of an initial set of data with a continuous improvement of the knowledge on the
applications analysed. The methodology requires two main operative phases with different know-how content.

References

[1] EN ISO 10218-2:2011

[2] ISO/TS 15066:2016
[3] EN ISO 10218-1:2011
[4] https://www.universal-robots.com/
[5] https://www.kuka.com/en-us/products/robotics-systems/industrial-robots/lbr-iiwa
[6] http://www.fanuc.eu/it/en/robots/robot-filter-page/collaborative-cr35ia
[7] http://www.comau.com/EN/media/news/2016/06/aura
[8] Teiwes et Al. “Identifying the Potential of Human-Robot Collaboration in Automotive Assembly Lines using a Standardised Work
Description”. In Automation and Computing (ICAC), 2016 22nd International Conference on, (2016). DOI: 10.1109/IConAC.2016.7604898
[9] Grahn, Langbeck. “benefits of collaborative robots in assembly – an evaluation scheme” In: The 6th Swedish Production Symposium; (2014).
http://conferences.chalmers.se/index.php/SPS/SPS14/paper/viewFile/1733/405
[10] Introduction to MTM-UAS, http://mtm-international.org/introduction-to-mtm-uas/