Euroc Proposal MTC Lu Uob Ab

Joint Robot-Human Logistics and
Assembly in Aerospace
Team: MTC-LU-UoB-Airbus
Document date:
9 February 2015
________________________________________________________
Project acronym:
Project full title:
Call identifier:
Grant agreement no.:
Starting date:
Duration in months:
Lead beneficiary:
Project web site:
EuRoC
European Robotics Challenges
FP7-2013-NMP-ICT-FoF
CP-IP 608849
1 January 2014
48
CREATE
www.euroc-project.eu
[confidential]
Annex I Proposal template
Page 1 of 26
Guide for proposal preparation
CP-IP 608849
CONTENTS
CONTENTS ............................................................................................................................................... 2
GENERAL INFORMATION ........................................................................................................................ 3
Team name ....................................................................................................................................... 3
Challenge option............................................................................................................................... 3
Challenge Team Leader (CTL) ........................................................................................................... 3
Participants (1 table for each organization, including that of CTL) .................................................. 3
Summary........................................................................................................................................... 5
SCIENTIFIC AND TECHNICAL QUALITY ..................................................................................................... 6
Proposal target .................................................................................. Error! Bookmark not defined.
Progress beyond the state-of-the-art............................................................................................... 9
S/T methodology and associated work plan .................................................................................... 9
IMPLEMENTATION ................................................................................................................................ 15
Individual participants .................................................................................................................... 15
Description of the partnership ....................................................................................................... 19
Overall Resources Costs and funding .......................................................................................... 20
IMPACT ........................................................................................................................................... 21
Expected results ............................................................................................................................. 21
Exploitation plan of results and management of knowledge and of intellectual property ........... 22
Ethical issues................................................................................................................................... 23
APPENDIX: Articles II.26, II.27, II.28 and II.29 Annex I of the EuRoC Grant Agreement .................... 24
[confidential]
Page 2 of 26
CP-IP 608849
GENERAL INFORMATION
Team name
MTC-LU-UOB-Airbus
Challenge option
Challenge 1 Reconfigurable Interactive Manufacturing Cell
Challenge 2 Shop Floor Logistics and Manipulation
Challenge 3 Plant Servicing and Inspection
Challenge Team Leader (CTL)

Name
Jeremy Hadall
Organization
Manufacturing Technology Centre
Phone Number
+44 2476701600
E-Mail
jeremy.hadall@the-mtc.org
Participants
Organization
name
Key person
Jeremy Hadall
Street Name
Pilot Way, Ansty Park
House Number
ZIP Code
CV7 9JU
Phone Number
+44 24760 701600
E-Mail
jeremy.hadall@the-mtc.org
Website
www.the-mtc.org
PIC (1)
959490342
City
Coventry
Country
United Kingdom
Role
Research Team
Organization
name
Loughborough University
Key person
Dr Laura Justham
Street Name
Wolfson School of MME
House Number
City
Loughborough
ZIP Code
LE11 3TU
Country
UK
Phone Number
01509 227553
E-Mail
L.Justham@lboro.ac.uk
Website
http://www.lboro.ac.uk/departments/mechman/staff/laura-justham.html
PIC (1)
999990752
Role
Research Team
[confidential]
Page 3 of 26
CP-IP 608849
Organization
name
University of Birmingham
Key person
Professor Jeremy L Wyatt
Street Name
School of Computer Science
House Number
Edgbaston
City
Birmingham
ZIP Code
B15 2TT
Country
UK
Phone Number
+44 121 414 4788
E-Mail
jlw@cs.bham.ac.uk
Website
www.cs.bham.ac.uk/~jlw
PIC (1)
999907526
Role
Research Team
Organization
name
Airbus
Dr. Ingo Krohne Technology Manager Automated Assembly
Key person
Dr. Robert Alexander Goehlich Manufacturing Engineering Project Manager

Dr. Hans Pohl Contract Manager
Street Name
Kreetslag
House Number
10
City
Hamburg
ZIP Code
21129
Country
Germany
Role
End User
+49 4074361667
Phone Number
+49 4074364054
+49 4074372822
ingo.krohne@airbus.com
E-Mail
robert.goehlich@airbus.com
hans.pohl@airbus.com
Website
www.airbus.com
PIC
999964756
The above information, together with the summary of the proposal, must be provided through the
web form available on http://www.euroc-project.eu. Instructions are available on the web. For
further information, please contact EuRoC at info@robotics-challenges.eu.
[confidential]
Page 4 of 26
CP-IP 608849
Summary
Human robot collaboration on the shop floor presents a significant set of challenges. In this project
we will develop methods for enabling a shop floor mobile robot to assist a human in logistical and
assembly tasks, by bringing tools, parts and assisting in the assembly process to reduce the strain
on the human worker. The scenario is as follows. In structure assembly, equipment installation and
final assembly of an aircraft, for example, many operations are repetitive, and can involve
manipulating quite heavy tools for long periods of time in sometimes un-ergonomic positions. In
this project we will develop a shop floor application of the EuRoC platform to enable safer, faster,
and more ergonomic operations on an airframe by using a robot in collaboration with a human.
The scenario is an assembly operation on the inside or outside of an airframe. The robot will pick
and carry a large number of fixing elements or assembly parts and the relevant tools from a
storage area and bring them to a worker at the airframe assembly. The robot and the human will
then work together to apply the fixing elements and parts. The key steps in the scenario are:
i)
ii)
iii)
iv)
v)
vi)
vii)
Identifying and grasping required tools and materials from a work-surface.

Placing the tools and materials on the mobile platform in a safe configuration for
transport.
Navigating to the worker, and waiting alongside.
Grasping materials or a tool taking into account wrenches to be applied in its
operation.
Positioning the material or presenting the tool for co-manipulation by the human, who
will then apply it to each fixing to perform the task.
The robot will learn the human applied forces, and assist the human by adapting the
end effector forces over time (i.e. in a gravity compensated mode)
The robot will interpret foot motion in order to follow the human as they walk along
the airframe to enable continued fixing.
This scenario covers a variety of important human-robot collaboration skills in a shop floor logistics
and manufacturing setting. Our contention is that to be useful in many practical settings it is as
important to allow the robot to be used as intelligent tool guided by the human, as it is for the
robot to operate fully autonomously, e.g., for picking and moving parts. This proposal is predicated
on the notion that to achieve fluid interaction much of the technical challenge lies in the robot
knowing when to switch from one mode to the next on the basis of interpreting the humans
intentions from sensory input. The key scientific and technical challenges that must be met which
go beyond the state of the art are:
i)
Flexible Grasping. Grasping and fetching of parts and tools of possible novel shape.
We will build on our recent breakthrough work on this.
ii)
Force based Human-Robot interaction. The robot works as a multiplier of the humans
own abilities, co-applying forces at the end effector in the desired way.
iii)
Adaptive Interaction and Machine Learning. Learning will be used to allow the human
to train the robot over time to apply assistive forces optimally. In addition the robot
must interpret the humans body motions to correctly follow the human.
[confidential]
Page 5 of 26
CP-IP 608849
SCIENTIFIC AND TECHNICAL QUALITY

Proposal Target. Challenge 2 is concerned with demonstrating in a real industrial
environment the benefits of mobile manipulation, with specific focus on the RTD issues
involved in establishing safe and effective human-robot collaboration, so that robot coworkers can operate effectively in an unstructured environment. The principal characteristic
of real shop floor environments is that the robot must work alongside humans who already
perform a huge variety of manipulation tasks. These include i) fetching tools and parts, ii)
carrying them to the work-station, ii) orientating parts relative to the main assembly, iv) part
mating, v) applying tools to parts, and vi) applying and manipulating deformable materials
such as glues or surface finishes. The human workers perform these tasks in a largely
unpredictable sequence, due to adaption to other workers, availability of parts and other
uncontrollable factors. Therefore, realistically very few of these tasks are fully autonomously
executable by a robot using todays state of the art. Most involve complex, context sensitive
decision-making as well as force-based dexterous manipulation that is far beyond the state
of the art in autonomous robotics.
One approach to this is to apply the robot only to the parts of the task it can perform fully
autonomously, and let the human perform the rest. But this is often inefficient. In addition
real companies problems are ones where logistics and assembly are closely intertwined,
where humans perform tasks with heavy tools in un-ergonomic positions, and it is typically
not the case that automating the simple pick and place operations alone brings much
economic and ergonomic benefit. Thus to ensure wide and early uptake of mobile
manipulation we need to create robots that can be used for a much wider range of purposes
than their purely autonomous abilities allow. In this team we place particular focus on using
the robot not just for logistics, but also to eliminate as much as possible the physically stressful
operations common in manufacture. This is an ideal scenario for the challenge hardware,
since it exploits force-controlled manipulation.
We will develop a robot co-worker that switches between different levels of autonomy as
required, with an emphasis on the human being able to use the robot as an intelligent, force
sensitive tool. We call this multi-level compliant autonomy. At one end the robot is fully
autonomous for operations such as pick and place. At the other end the robot has no
autonomy and is merely a compliant, gravity compensated tool and/or material holder held
by the human, entirely under their direct manual control and achieving ergonomic benefit
simply by removing the weight of the tool. In between it uses learning to act as an adaptive
force-multiplying tool, learning the forces applied by the human to perform a task and then
replicating those as closely as possible allowing the human to achieve maximum ergonomic
benefit by minimising the forces they apply to achieve a task. In this way the robot can also
acquire a statistically significant number of force profiles for each task. This will in turn allow
future versions of the robot to learn to autonomously perform the task after receiving
sufficient training, and as its perceptual abilities increase.
Thus the approach is intended to create impact in industry by using the EuRoC hardware in a
way that obtains the maximum economic benefit. Only in this way we will encourage
significant investment by companies, by increasing the return on investment (ROI) beyond
what pure autonomy would allow. The creation of such a royal road to adoption of mobile
manipulation on the shop floor is a critical goal of EuRoC.
[confidential]
Page 6 of 26
CP-IP 608849
Figure 1: Human workers perform hours of repetitive fixing/riveting operations within the
airframe in difficult access areas.. Examples of the airframe before and after panel fixing are
given above. (Source: Airbus)
Our specific objectives to achieve this vision are:
O1 Flexible Autonomous Grasping. We will enable grasping of parts and tools of novel shape.
We will build on our recent breakthrough work on this. [T1 Freestyle]
O2 Force based Human-Robot interaction. We will develop methods where robot works as
a multiplier of the humans own abilities, co-applying forces at the end effector in the desired
way and perform this in a safe manner. [T2, T3]
O3 Adaptive Force Based Interaction. Learning will be used to allow the human to train the
robot over time to apply assistive forces optimally. In addition the robot must interpret the
humans body motions to correctly follow the human in a safe way, using force signals as
feedback from the human to the robot. [T2 T3]
O4 Requirements, acceptance and dissemination. To ensure the appropriateness of the
technology we will spend considerable effort in understanding the ways that human workers
currently operate tools to carry out their tasks. We will then identify how robot collaboration
will be best suited to support the human worker for each task. [T3]
Stage II: Showcase Round: End User Driven Task
This intermediate task will demonstrate a robot working alongside a human using our forcebased interaction approach to perform a task. The task will require the robot to perform
initially in gravity compensation mode, learning the forces applied by the human, and
adjusting the forces it applies so that the human can achieve easier working. The human will
grasp the robot arm so as to guide it, pushing the tool into place and sensing the forces
applied. As the human operates, the robot will build a force-based model of the activity, by
measuring the reaction forces applied at its joints and the tool centre point. In the next stage
the robot will then adapt the forces it applies during the task to allow the human to
accomplish the task while applying less force. Critical questions here are how to adapt the
applied force, and how the human adapts to this change in robot behaviour. We will show a
system that enables the human to perform a simple task, such as riveting, for a specific
application. This task will be complementary to the demonstration in the free-styling round,
where we will develop and demonstrate our autonomous dextrous grasping technology.
Stage III: Field Tests: The Use Case
Motivation: The end-user is Airbus, and their task is concerned with airframe assembly.
Airframe assembly has several properties that have driven the formulation of our technical
approach. First, many of the workers have to lift tools for extended periods of time. Even with
quite light tools this can cause ergonomic strain. In addition, some of the operations
[confidential]
Page 7 of 26
CP-IP 608849
Figure 2: Two examples of the fixings to be applied: the Hi-Lite collar and Hi-lock pin, and on
the right a typical tool that must be aligned around the fixing element.
are in un-ergonomic positions. For example they might involve hundreds of riveting or fixing
operations above the head of the worker for a long stretch of time. Alternatively it might
involve applying tools within a physically highly constrained space. It is of considerable and
direct economic and ergonomic benefit to Airbus to begin to automate these tasks, but we
believe it is not yet possible or desirable to fully automate them. Hence this is the motivation
for the multi-level autonomy approach. In addition Airbus are clear that their shop floor does
not benefit by merely automating the logistics aspects, but by combining this with an
approach to human robot co-working. The specific stages in the Use Case Scenario are as
follows.
i)
The robot travels to a tool/part station, grasping the tools and bags of fixings.
ii)
The robot navigates to the worker and waits until the task start is indicated.
iii)
It grasps a tool offered by the human, taking into account the typical wrenches to
be applied in its operation.
iv)
The robot will present the tool for co-manipulation by the human, entering a
gravity compensation mode. The human will then apply it to perform the task.
v)
The human will guide the robot to the lock pin in order for the robot to perform the task.
Machine learning will take place, considering the human applied forces and trajectories,
to enable the robot to develop the capability to become autonomous.
S&T Issues. This proposal contains a number of challenges with regards to automation and
human-robot collaboration within the manufacturing domain. This is particularly pertinent
within high value manufacturing, such as the aerospace industry, where small batches of
products are manufactured, resulting in a highly labour intensive process. The main technical
challenges identified within this proposal are:
1 Autonomous Grasping. The robot must grasp tools and parts reliably. It must identify the
correct object, and grasp tools in the face of object shape variation.
2 Human Sensitive movement. The robot must navigate through the shop floor, locate,
meet and then follow the human as they work. This requires movement interpretation.
3 Learning and Assistance in Joint Human-Robot Tool Use. Machine learning of human tool
use is a challenging problem, particularly when the trajectories and forces vary, or when the
forces change suddenly during contact between the tool and the work-piece. Correct
application of assistive forces is also an S&T challenge.
4 Multi-level Autonomy. This will be handled using a behaviour based approach to switch
between autonomy settings based on triggers from the human co-worker. These will include
the ability to stop the robot and to indicate the task stage.
[confidential]
Page 8 of 26
CP-IP 608849
Progress beyond the state-of-the-art

Grasping. In this project we will build on machine learning methods for grasping to achieve
mobile manipulation with a dexterous hand. The key is to be able to learn dexterous grasps
that generalise to new objects. Previous work utilises common object (Saxena et al., 2008;
Detry et al., 2013; Herzog et al., 2014; Kroemer et al., 2012). This works well for non-dexterous
hands. Recently we (Kopicki et al 2014) showed how to generalize grasping across object
categories with dexterous hands. Our technique was able to grasp 45 unseen objects from 5
example grasps. Here we will transfer it to the Shadow Hand, using a standard IK solver to
place the robot base in a position where the object is graspable. The method relies on a depth
cloud. We will obtain one from the depth sensor on the challenge platform.
Learning and Assistance in Joint Human-Robot Tool Use. The idea of a human directly
manipulating a compliant robot arm is a long-standing one. The determination of forces
during a contact interaction between a human and a Kuka LWR robot has been shown, and
also how the robot can modify the control law to make safe contact forces between the
human and the robot (Magrini 2014). In this work we will exploit this scheme to determine
the contact forces. We will go beyond this to learn the contact forces at the tool centre point
during application of a tool. This requires separation of the contact forces exerted by the
human from those at the tool contact. We achieve this by combining the above method with
an additional inertial tensor estimate from a force-torque sensor at the robot wrist.
Aerospace Automation: is still largely dominated by human operators assembling parts and
products by hand. This results in increased variability in product quality (Jamshidi 2010). There
is R&D to increase the level of automation within the domain, e.g. within wing (Irving 2014)
and engine manufacture (Tingelstad 2014). However, due to the size of the component parts
being manufactured, the associated difficulties with manipulating a robot into hard to reach
areas and the risks associated with robots working overhead, their uptake has not been
widespread or wholly successful. We will move beyond this by building the mobile platform
into a human governed process.
S/T methodology and associated work plan

Overall Strategy
We have split out the autonomous and joint human robot aspects of this project across Task
T1 (Freestyle) and T2 (Showcase) respectively. In T3 we will integrate the solutions to each
part and thoroughly field test at Airbus in Hamburg. The sub-tasks within these are organised
as described in detail in the Task tables.
Project Management The research partners will work closely with Airbus, to ensure that the
solution adds value to their manufacturing process. We will have regular remote meetings to
monitor project progress, with the team leader ensuring the work-plan is being followed. We
will use version control (GitHub) to develop the best possible code-base. We will hold two
face-to-face challenge team meetings during stage II to coincide with internal presentations
and key deliverables and milestones i.e. at month 10 and month 15
[confidential]
Page 9 of 26
CP-IP 608849
Table 1a: Task list

Task
no.
Task title
Lead partic.
short name
Personmonths
Start
month
End
month
T1
Round (a) of Stage II: Benchmarking+free-style

round
UoB
37
Mo1
Mo10
T2
Round (b) of Stage II: Showcase round
LU
20
Mo11
Mo15
T3
Stage III (Field tests Pilot Experiments)
MTC
36
Mo18
Mo26
TOTAL
93
Tables 1b: Brief description of tasks

Task no.
T1
Start month
Task title
Round (a) of Stage II: benchmarking+free-style (1)
Participant no.
Participant short
name
Person-months per
participant
Mo1
End month
MTC
LU
UoB
Airbus
12
12
12
Mo10
Objectives
Solve the usecase driven task defined by the challenge host (benchmarking round)
Solve a task chosen by the team (freestyle round)
Description of work and role of participants

Here we describe the free-style task chosen by the team. This is a tool grasp and carry task. The platform will
grasp the necessary tools and bags of parts sitting on a work-surface and navigate with them to a human
worker on the shop floor. We will work with the Challenge hosts to enable the use of a dextrous manipulator
(Shadow Hand) to be integrated with the challenge platform. This dextrous manipulator will subsequently be
employed in our End Use Case to allow a firm grasp of tools designed for human hands. For the free-style task
we will integrate this hand with our dexterous grasping software (UoB) for novel objects. We will devote 18
PMs to the Benchmarking tasks and 18 PMs to the FreeStyle task. The specific tasks steps are as follows:
Subtask 1.1 Integration of Shadow Hand with Challenge Platform. This has been raised with the EuRoC
consortium and it has been agreed that this should be feasible for the free-styling rounds. The consortium, led
by MTC will work with DLR to integrate our Shadow Hand with the Challenge platform. (4PMs) (MTC)
Subtask 1.2 UoB grasping software integration with Shadow Hand. The UoB grasping software (GRASP and
GOLEM), are integrated with PCL, and functional on the DLR-HIT hand 2, the Schunk dextrous hand, and the
Justin three fingered hand. We will integrate this software with the Shadow Hand, writing the necessary
device drivers or interfaces using the ROS stack. The hand will be fully tested at UoB and MTC. The drivers for
the hand will allow compliant control, using a control framework developed at UoB. (5PMs) (UoB)
Subtask 1.3 Integration of Grasping software with Challenge Platform Perception. The perceptual
component of the grasping software is a PCL based library that uses point clouds to represent the object to be
grasped. We will integrate this with the challenge platform stereo camera. (5PMs) (UoB)
[confidential]
Page 10 of 26
CP-IP 608849
Subtask 1.4 Integration and Testing of Grasping software with Mobile motion planning. A planner is required
that coordinates the base pose with the grasping planner for the arm/hand. We will write an interface
between the two that achieves a feasible position next to the tool and part station. (4PMs) (LU, MTC)
Subtask 1.5 Grasp Testing. We will test the grasping abilities of the system based on the software from
Subtasks 1.1-1.4. We will use a force-based assessment of grasp quality, and plan a re-grasp if the grasp is not
promising. Tools will be presented upright to afford grasps. (5PMs) (UoB, MTC)
Subtask 1.6 Integration of Person Search. We will develop a suitable person detection routine, test and
integration on the challenge platform so that it can identify and approach workers. (5PMs) (LU, MTC)
Subtask 1.7 Whole system testing. We will system test on a variety of tools from our end-user scenario, and
show the ability to grasp these and the parts bag and take them to a human co-worker. (5PMs) (MTC, LU, UoB)
Risk Management: We retain 4 PMs of effort to allocate as necessary as a time contingency.
Roles: In the Freestyle task UoB will provide expertise in grasping software, and also in point cloud processing.
LU and MTC will develop the mobile motion planning to support grasping and person search. The integration
between the Shadow Hand and the challenge platform will be handled by MTC in conjunction with the
Challenge Hosts. All partners will contribute to system testing.
Task no.
T2
Task title
Round (b) of Stage II: showcase round (1)
Participant no.
Participant short name
Person-months per
participant
Start month
Mo11
MTC
LU
UoB
Airbus
End month
Mo15
Objectives
Solve the enduser driven task aimed at showcasing customizability under realistic manufacturing
conditions

During this task the precise end user requirements for T3 will be captured. In addition, the joint human-robot
tool use will be showcased on a demonstrator at the MTC. The tool will be held by the robot but guided by the
human to perform a task, such as riveting or drilling, where the tools have to be held, in many cases, in unergonomic positions by the human for long periods of time. The robot will hold the tool in a known orientation,
such that the human may guide it to the lock-pin position, ready to perform the task. The robot will initially
operate in a passive gravity compensation mode, but subsequently will apply assistive forces to support the
human operator during task execution. Both the robot and human will have clear, complementary roles to
improve the quality and speed of the combined process.
Subtask 1.1 Implementation of end user requirements into platform design
The consortium will work with the end user to ensure the demonstration airframe is accurate with respect to the
current as-is scenario at Airbus. The platform and demonstration airframe will be commissioned, and a task
analysis performed on the drilling and riveting tasks as they are currently performed by humans, and also for a
panel placement task that will act as a fall-back. Any changes required to ensure safe co-working will be identified.
We will thus characterise the requirements on the robot platform for the end user driven task (2PMs) (LU, UoB,
MTC, Airbus).
Subtask 1.2 Development of Gravity compensated tool support mode
[confidential]
Page 11 of 26
CP-IP 608849
The robot platform must be capable of grasping the tool and ensuring that it is held firmly for operation. This
will be achieved using grasp stabilisation software for the Shadow hand currently in the ROS stack for the hand.
The next step is to achieve gravity compensation given the additional mass near the end effector. To achieve
this the external forces imposed by the motion of the arm while holding the tool will be estimated by moving
the tool through calibration motions performed offline for a fixed set of tools. The human will then execute the
task (drilling or riveting) using the robot in this gravity compensated mode. (2PMs) (LU, MTC, UoB)
Subtask 1.3 Learning of force characteristics of tool use
Having performed in gravity compensation mode, the robot platform can learn the forces applied by the human
for various tasks and manufacturing scenarios. This will be done by modelling the force-based activity of the
human; measuring the reaction forces applied at the robot joints and at the tool centre point. The machine
learning approach we will apply will learn the parameters of movement primitives describing the motion, and
also the sequence of forces applied. The goal of this subtask is to teach the robot the characteristic force curves
for each task to enable it to be more adaptive for sub-task 1.4. (8PMs) (LU, MTC, UoB)
Subtask 1.4 Force assisted tool use
The robot will be able to adaptively sense the mode of operation and adjust its own force profile such that the
human experiences easier working conditions. We examine the effectiveness of predictive and feedback
control to create the best experience of working with the tool. We will investigate how the human adapts to
the robot assisted manufacturing process. Throughout this task, the human will remain the skilled operator
required to make decisions about the positioning of the robot TCP. (6PMs) (LU, MTC, UoB)
Risk Management: We retain 2PMs of effort across the project partners as a time contingency.
Roles: In this round LU will provide expertise in the user requirements definition, machine learning and
platform development. MTC will lead system integration and develop a test facility. UoB will contribute
expertise in control and learning. Airbus will lead the task analysis with MTC.
Task no.
T3
Task title
Stage III (Field tests Pilot Experiments) (1)
Participant no.
Participant short name
Start month
Mo18
MTC
LU
UoB
Airbus
11
11
11
Person-months per
participant
End month
Mo26
Objectives
Provide an initial solution to the Use Case task by sequencing the solutions developed in the Freestyle
and Showcase Rounds.
Assess the assistive tool use idea and acceptance testing with workers.
System integration and on site proof of concept.

In this task we will focus on sequencing and integrating all the elements from the first two stages. In particular
the freestyle round will provide the grasping technology for the logistics aspects. Second the assistive tool use
will come from the Showcase round. In this round we will thoroughly acceptance test the system in the Airbus
plant in Hamburg. This will also involve integration and systems testing of the whole system in situ.
Subtask 1.1 - Systems Integration and Testing (Factory Acceptance Test)
The consortium (led by MTC) will work with the end user and benchmark solution provider to integrate the
technologies developed through the freestyle and showcases rounds into one platform ready for testing at
[confidential]
Page 12 of 26
CP-IP 608849
Airbus Hamburg. Systems integration and testing will be carried out on the benchmark solution against a
factory acceptance testing schedule developed by the partners to fully test systems integration and operation.
(8PMs) (MTC, LU, UoB)
Subtask 1.2 - Field Trials Of Integrated Systems
Once the developed system has been fully tested at MTC, it will be transported to Airbus Hamburg for field
testing. Working with the end-user, the challenger team will develop a test schedule and process to prove the
developed solution in the end users environment. In particular these tests will focus on proving the grasping
and assistive tool concepts. At this stage, testing will be carried out with the consortiums researchers to ensure
safe operation. (8PMS) (MTC, LU, UoB, Airbus)
Subtask 1.3 - Acceptance Testing and System Adjustment
Following the initial field trials (and remedial actions) a full acceptance test will be will be carried out. These
tests will be carried out to a prescribed Acceptance Test Schedule giving confidence in the systems capability to
a high level. Should adjustments be required, these will be assessed for impact on the overall system and either
re-tested or a concession made for that particular part of the test. (8PMs) (MTC, LU, UoB)
Subtask 1.4 - Field Testing with Workers and Handover to Airbus
Following satisfactory passing of the acceptance test, workers from Airbus will be trained in the use of the
systems. The challenger team will remain on site during field-testing and handover to ensure a smooth
transition. To ensure that a full transition is made and to assess the impact of the technology, a transition
handover plan and check list will be developed. This will be reviewed over a period of two months to highlight
any issues or failings that may be addressed by the Challenger team whilst on site. The partners will develop an
assessment method to measure impact beyond the duration of EuRoC. (8PMs) (MTC, LU, UoB, Airbus)
Risk Management: As with other tasks, we retain 4PMs of effort to allocate as a time contingency.
Roles: In the Field Testing task, MTC will provide systems integration and factory testing capabilities, experience
and a test environment. LU and UoB will develop test procedures (with input from MTC and Airbus) and will
provide field testing operators. Airbus will provide a field test environment in their Hamburg facility and in Task
1.4, production operatives to carry out field testing of the developed solution. All partners will contribute to
system testing and will collaborate to develop a long term impact assessment process.
Table 1c: List of deliverables

Deliverable Deliverable name
no.
Task(s) Nature Dissemination

involved
level
Delivery
date
D1
Report describing the activities of the team and

achieved results in Round (a) of Stage II
T1
R (1)
PU
Mo10
D2
Report describing the activities of the team and the

achieved results in Round (b) of Stage II
T2
RE
Mo15
D3

achieved results in Stage III
T3
RE
Mo26
Table 1e: List of milestones

Milestone
no.
Ms1
Milestone name
Round (a) of Stage II
completed
Task(s) Expected
Means of verification
involved
date
T1
Mo10
Report describing the activities of the team and

achieved results in Round (a) of Stage II completed.
Mid-term evaluation successfully passed.
[confidential]
Page 13 of 26
CP-IP 608849
Ms2
Round (b) of Stage II

completed
T2
Mo15

achieved results in Round (b) of Stage II completed.
End-user driven task successfully solved.
Ms3
Round III completed
T3
Mo26

achieved results in Stage III completed.
Final Challenge Workshop attended.
Table 1f: Summary of staff effort

Partic. no. Partic. short name Task 1
Task
2
Task3 Total person months
MTC
12
11
29
LU
12
11
29
UoB
12
11
29
Airbus
Total
37
20
36
93
Table 1g: Risk assessment and contingency plan

Risks
Probability
Impact
Remedial actions
Force assistance is unnatural

with predictive method
Medium
High
ORevert to gravity compensate or only apply

additional forces during simple, straight line, higher
force moves.
Difficult access areas also

too challenging for robot
Medium
Low
Use only robot for areas that it can reach to

demonstrate initial capability, e.g. externally
Robot arm cannot reach

selected positions
Low
High
Use only positions that are in reach with robot arm
Shdow hand cant be used

with the Challenge platform
Medium
High
We have access to a lightweight Pisa-IIT hand that

is also strong and can grip the tool well.
Robot is too slow to be of

economic use to Airbus
Medium
High
Investigate lighter weight platform post EuRoC.

EuRoC only provides proof of concept
Riveting use case is too

difficult for robot
Medium
High
Fall back to drilling, or panel placement (for which

we would use suction cups not a hand)
Member of the consortium

drops out of the challenge
Low
Medium
The other members will be able to pick up the

work.
Hardware unable to meet

requirements of end user
Medium
Medium
Relax requirements for proof of concept. Search for

better follow on hardware.
Health & safety prevents

field testing
Low
Medium
Safety risk assessments will be carried out in

advance of field tests to reduce this likelihood
[confidential]
Page 14 of 26
CP-IP 608849
IMPLEMENTATION
Individual participants
Table 2a: Team participants
Participant 1

(MTC)
Brief description of the As one of the founding members of the High Value Manufacturing Catapult,
organisation
the Manufacturing Technology Centres aim is to develop and promote
world class manufacturing processes and techniques for UK industries.

Working from a state of the art research facility its expertise in intelligent
automation and manufacturing systems will be invaluable in solving the
significant problems of creating a flexible manufacturing system.
www.the-mtc.org
Role in the project and
main tasks
MTC is the lead partner in the consortium, and will provide systems
integration, expertise in aerospace assembly, and lead T3 on Field Tests.
Relevant previous
experience
The MTCs Intelligent Automation team draws on over 200 person years of
industrial automation experience across a wide range of industries. Recent
examples are the lead demonstration of the FP7 LOCOMACHS project,
advanced industrial robotic controls (including ROS-Industrial), metrology
assisted robotics, and the application of robotics to complex fabrications
and assembly tasks. MTC has over 5 million of industrial scale automation
hardware for development and demonstration work, including the Shadow
Robot hand that will be integrated on the test platform for this challenge.
Profile of team
members that will
undertake the work
Jeremy Hadall - holds a Masters in Engineering and has 15 years experience

in developing innovative manufacturing systems, particularly advanced
automation and robotics. He has worked with: ABB, Tesco, Rolls-Royce,
Land Rover, BMW and has significant experience in collaborative projects.
He is chair of the UKs High Value Manufacturing Catapult Automation
Forum, sits on the Board of the EPSRC Centre for Innovative Manufacturing,
and is Chief Technologist for Automation at the MTC.
Richard Kingston holds a BEng in Automotive Engineering and 20 years
experience in industrial robotics. He has worked on many robotics projects
within the automotive and aerospace sectors for companies throughout
Europe and the USA. For the last 15 years he has worked on metrology and
software development, especially in relation to robotic applications. He has
previously worked on the COMET FP7 project within a software
development role providing integration between robot and metrology
systems. He is currently a Technical Specialist (Robotics) within the
Intelligent Automation area theme at Manufacturing Technology Centre.
[confidential]
Page 15 of 26
CP-IP 608849
Participant 2
(LU)
Brief description of
the organisation
Loughborough University (LU) www.lboro.ac.uk is a leading UK researchintensive university. The Wolfson School of Mechanical and Manufacturing
Engineering (WSMME) holds two prestigious Queen's Anniversary Prizes in
High Value Manufacturing, and is ranked 2nd in the UK. The EPSRC Centre for
Innovative Manufacturing in Intelligent Automation (IMIA) is the EPSRC
national centre for intelligent automation research.
Role in the project

and main tasks
LU will lead Task 2 and support the other two tasks. The EPSRC Centre at LU
has an established stream of research on collaborative automation, which
aims to enhance the use of existing worker skills via robot collaboration.
Relevant previous
experience
The EPSRC Centre is a research collaboration between Industry, Academia and

the MTC. The principle industrial partners of the centre are from the
Aerospace industry and therefore the team at LU have a strong track record
working in this sector on numerous projects and challenges. Aerospace has a
number of specific requirements, which are not prevalent within other
industries, and the team at LU have extensive expertise working alongside
these industrial partners. In addition, the state of the art laboratory based at
LU has extensive industrial robotics equipment, both hardware and software,
which can be utilised in support of the challenge.
Profile of team
members that will
undertake the work
All four members of the LU team are members of the EPSRC Centre for
Innovative Manufacturing in Intelligent Automation.
Dr. Laura Justham is a member of the EPSRC Centres academic team and
has been working on human robot collaboration, mobile robots in aerospace
applications, and industrial computer vision systems for over 10 years.
Dr. Niels Lohse is a senior member of the EPSRC Centre and has extensive
experience working on research projects, including projects in the aerospace
sector with Airbus and Rolls-Royce. Niels research interests include
manufacturing system modelling, human-machine interaction, machinelearning, and adaptive systems.
Dr. Zahid Usman is a research associate with over 7 years experience
working on modelling and simulation within a manufacturing environment.
Dr. Phil Ogun is a research associate with experience of machine vision tasks
within the manufacturing domain and the integration of intelligently
automated solutions.
[confidential]
Page 16 of 26
Participant 3
CP-IP 608849
University of Birmingham
(UoB)
Brief description of the The University of Birmingham (www.bham.ac.uk) is one the largest
organisation
research intensive universities in the UK. The School of Computer Science is
ranked 8th in the UK (REF 2014). The Intelligent Robotics Laboratory

(www.cs.bham.ac.uk/go/irlab) is a leading European lab in robotics and
artificial intelligence. The IRLab has expertise in robot vision (Leonardis),
task planning (Hawes, Wyatt), motor learning and multi-contact control
(Mistry), software architectures for robotics (Hawes, Wyatt), manipulation
(Wyatt), and machine and robot learning (Mistry, Wyatt, Hawes, Leonardis).
The IRLab will draw on its experience in building systems that perform
dextrous grasping, motor learning and human robot interaction (Burbridge,
Hawes, Zito, Wyatt). The lab has a substantial code base for mobility and
manipulation, which is used across multiple projects, including several
pieces of software for integration (CAST, GOLEM) that are used beyond
Birmingham. The laboratory has significant experience in working with
ROS, with designing integration frameworks linked with ROS, and in leading
projects involving large-scale software integration for autonomy.
Role in the project and
main tasks
UoB will lead T1, and in particular will lead the work on autonomous
grasping. Birmingham will also contribute to the work on learning of
trajectories and force-torque patterns for tool use.
Relevant previous
experience
UoB has significant experience in robotics projects having coordinated

three European projects (FP6: CoSy; FP7 CogX, GeRT, PacMan, CodyCo;
H2020: Cogimon). The relevant experience we will bring to this project is
our experience of grasping (GeRT, PacMan) including results from the Boris
manipulation system able to grasp novel objects from one-shot learning. In
addition we will bring experience in force-based manipulation from
PacMan, CodyCo and Cogimon, with an emphasis on learning interactions.
Profile of team
members that will
undertake the work
Professor Jeremy L Wyatt has published more than 90 papers on robot task
planning, motion planning, robot learning, manipulation, vision and
machine learning. He has won two best paper awards, has career funding of
10m, coordinated major EU projects such as PacMan and CogX, and
participated in three more (CoSy, GeRT, Strands).
Dr Nick Hawes works on the application of AI to robots that can work with
humans, architectures for intelligent systems, AI task planning for robotics,
and qualitative spatial representations for robots. He worked on both the
CogX and CoSy projects. He coordinates the FP7 STRANDS project on
robot-human interaction.
Dr Chris Burbridge is a research fellow with expertise in mobile robotics
robot manipulation and task and motion planning. He has worked on the
successful FP7 GeRT project. He will lead implementation.
Dr Marek Kopicki is a research fellow working on dexterous robot
manipulation. He will provide expertise in dextrous manipulation.
Mr Maxime Adjigble is a research engineer with expertise in force
controlled exoskeletons and control engineering for robotics.
[confidential]
Page 17 of 26
CP-IP 608849
Participant 4
Airbus
(Airbus)
Brief description of
the organisation
Airbus is a leading aircraft manufacturer that consistently captures around

half of all orders for airliners with more than 100 seats. Our product line-up,
which covers a full spectrum of four aircraft families from a 100-seat singleaisle to the largest civil airliner ever, the double-deck A380, defines the
scope of our core business.
Our mission is to be a top-performing enterprise making the best aircraft
through innovation, integration, internationalisation and engagement.
Based in Toulouse, France, Airbus is a division of the Airbus Group. It
employs directly some 55,000 people of over 100 nationalities. Airbus'
design and production sites are grouped into four wholly-owned
subsidiaries, Airbus Operations SAS, Airbus Operations GmbH, Airbus
Operations S.L. and Airbus Operations Ltd. *
* www.airbus.com
Role in the project

and main tasks
Airbus is the End User partner in this team consortium and will provide
expertise in assembly in aerospace.
Relevant previous
experience
Various R&T experiences in the scope of automation strategies for

aerospace applications.
Profile of team
members that will
undertake the work
Dr. Ingo Krohne works in the Manufacturing Engineering R&T Generic

Technologies department at Airbus in Hamburg plant.
Dr. Robert Goehlich works in the Manufacturing Engineering Process
Development & Transnational Harmonisation department at Airbus in
Hamburg plant.
[confidential]
Page 18 of 26
CP-IP 608849
Description of the partnership

The challenge team brings together four partners with complementary expertise. On the one hand we
have Airbus the end user, with world leading expertise in their domain of aircraft manufacture, and in
particular the possibilities and pitfalls of automation in the manufacturing process. They have a clear
view of the economic benefits of automation and the adoption of robotics, and will be the main
exploiter of the technology we will develop. At the other end we have the University partners: LU and
UoB, who provide the research base from which this project will grow. UoB have expertise in pure
autonomous robotics: manipulation, navigation, machine learning, and machine vision. They also have
long experience in running and leading robotics projects, and in producing research robot systems
that bring together many different technology elements. In addition they have significant expertise in
long-life reliable robots (e.g. Strands project) with experience in building robot systems that work in
the wild for periods measured in weeks. We will exploit their expertise in dextrous manipulation and
machine learning in particular. LU also have research expertise in intelligent robotics, but more
embedded in the domain of manufacturing: and so bring not just the basic research skills, but also the
expertise in applying these novel control strategies, machine learning, vision etc within manufacturing
settings. In addition LU have considerable experience of automation in the aerospace domain already.
Finally MTC work in the space of TRL 4-8, bringing research ideas along the TRL scale through feasibility
study to pre-product prototypes. Their experience is thus very much in the translational stage from
research to business. MTC have several team members with a total of more than 40 years of
experience in industrial robotics and advanced automation. MTC already work with Airbus, and have
extensive experience in projects for automation in aerospace, being the leader of one of the robotics
demonstrations in the FP7 project LOCOMACHS, which is precisely concerned with automation of the
assembly of structures in aeroplane manufacture. If we consider it in terms of the objectives we have
set then the match between objectives and partners is as follows:
O1 Flexible Autonomous Grasping. [T1 Freestyle] Here the role of Birmingham will be critical, as
they bring expertise in dextrous grasping of novel objects and objects under pose uncertainty. This
approach will be transferred to the Shadow Robot Hand, working with MTC who will bring expertise
in its use. LU and MTC will bring experience in systems integration and machine vision. UoBs expertise
in point cloud analysis will complement the stereo-vision expertise at MTC and LU.
O2 Force based Human-Robot interaction & O3 Adaptive Force Based Interaction. [T2 T3] To meet
these two objectives we will draw upon the expertise of LU in control and interaction, and on the
experience of UoB in machine learning. MTC will bring expertise in systems integration and Airbus will
specify the requirements for the interaction and the task.
O4 Requirements, acceptance and dissemination. [T3] The expertise of Airbus in the requirements
and case for robotic assistance will be crucial. LU will use their end user requirements definition
expertise to analyse the solution requirements. In addition Airbus and MTC will lead the acceptance
testing process. LU & UoB will contribute to dissemination via the print and broadcast media, web and
academic papers. Airbus and MTC will lead the dissemination process within Airbus, and direct the
exploitation efforts.
The three research partners are geographically close, LU and UoB being founding research partners in
the MTC. This means that they are already working together on a number of projects in intelligent
automation and robotics, so that the teams are well integrated. In the first stage of the EuRoC
competition MTC, LU and UoB worked together to produce our entry for the simulation competition.
This was successful, ranking us 4th out of the 17 teams that submitted an entry. This shows not only
that the partners have worked together, but that we know how to work together successfully within
the context of the joint development required for success in EuRoC.
[confidential]
Page 19 of 26
CP-IP 608849
Overall Resources Costs and funding

The team partners will mobilise and commit the required resources to guarantee the success of the
project. The total budget is 807k, and the requested funding 585k. The main costs are for the
personnel at the partner sites. In each case we have allocated 1 researcher to work full-time on the
project at each research site, giving 26 PM per site for MTC, LU and UoB. This gives us a steady 3
people to work on the project. In addition a small amount of time from senior researchers leading the
project has been set, 3 PMs per research partner. In addition Airbus will devote a total of 6 PMs to the
project. This brings us to a total of 399,618 for personnel. The other costs are as follows. We will
spend 27,584 on direct costs to support the project, notably the costs of integrating the Shadow
Hand with the challenge platform, producing the test rig for the showcase demonstration, computers
for software development, a small quantity of consumables for technical meetings, adaptation costs
to allow the fitting of the Shadow Hand to the platform, further small sensors, and shipping and
insurance costs for equipment to and from the partner site. Airbus will have, as part of this figure,
1000 for example parts and tools to be used by the robots. In addition to requested contribution
from the EC we will bring a Shadow Hand, a Kuka omnibot, and a Kuka LWR to the project as an
additional project contribution at no cost. We have costed travel at rates of 100/day for
accommodation, 200 per flight, and 50/day for subsistence costs. Travel thus totals 58,700 for the
entire duration of the project, including all challenge team meetings and exchanges.
Costs for Stage II
Partner MTC
Cost category
Personnel costs ()
Travel ()
Other direct costs ()
Audit certificates ()
Indirect costs ()
Total budget ()
Requested funding ()
Costs for Stage III

Partner MTC
Cost
43056
20000
(1)
(2)
(3)
(4)
(5)
20584
0
58126
141766
106324
Costs for Stage II

Partner LU
Cost category
Personnel costs ()
Travel ()
Indirect costs ()
Total budget ()
Cost
10000
Costs for Stage II

Partner UoB
Cost category
Personnel costs ()
Travel ()
(1)
Audit certifcates ()
(2)
Indirect costs ()
(3)
Total budget ()
(4)
(5)
Cost
32517
4700
(1)
(2)
(3)
(4)
(5)
539
0
43898
81654
61240
Costs for Stage III

Partner LU
85253
(1)
(2)
(3)
(4)
(5)
Cost category
Personnel costs ()
Travel ()
Indirect costs ()
Total budget ()
Costs (Stage II + Stage III)

Partner MTC
1000
0
57752
154005
115504
Cost
94277
10000
4500
0
56266
174043
130532
Cost category
Personnel costs ()
Travel ()
Indirect costs ()
Total budget ()
Cost
4000
Costs for Stage III

Partner UoB
Cost category
Personnel costs ()
Travel ()
(1)
(2)
Indirect costs ()
(3)
Total budget ()
(4)
(5)
Cost
75573
24700
(1)
(2)
(3)
(4)
(5)
20584
0
102024
223424
167564

Partner LU
49983
(1)
(2)
(3)
(4)
(5)
Cost category
Personnel costs ()
Travel ()
Indirect costs ()
Total budget ()
500
0
32690
87173
65380
Cost
49983
4000
0
0
32390
86373
64780
Cost category
Personnel costs ()
Travel ()
Indirect costs ()
Total budget ()
Cost
135236
14000
(1)
(2)
(3)
(4)
(5)

Partner UoB
Cost category
Personnel costs ()
Travel ()
(1)
(2)
Indirect costs ()
(3)
Total budget ()
(4)
(5)
1500
0
90442
241178
180884
Cost
144260
14000
4500
0
97656
260416
195312
[confidential]
Page 20 of 26
Costs for Stage II

Partner Airbus
Cost category
Personnel costs ()
Travel ()
(1)
(2)
Indirect costs ()
(3)
Total budget ()
(4)
(5)
Total costs for Stage II
Cost category
Personnel costs ()
Travel ()
(1)
(2)
Indirect costs ()
(3)
Total budget ()
(4)
(5)
Cost
24300
4000
0
0
16980
45280
22640
Cost
246886
44000
26084
0
198124
515094
375000
Costs for Stage III

Partner Airbus
Cost category
Personnel costs ()
Travel ()
(1)
(2)
Indirect costs ()
(3)
Total budget ()
(4)
(5)
Total costs for Stage III
Cost category
Personnel costs ()
Travel ()
(1)
(2)
Indirect costs ()
(3)
Total budget ()
(4)
(5)
Cost
20250
2000
1000
0
13950
37200
18600
Cost
152733
14700
2039
0
122928
292400
210000
CP-IP 608849

Partner Airbus
Cost category
Personnel costs ()
Travel ()
(1)
(2)
Indirect costs ()
(3)
Total budget ()
(4)
(5)
Cost
44550
6000
1000
0
30930
82480
41240
Total costs (Stage II+ Stage III)

Cost category
Cost
399619
Personnel costs ()
58700
Travel ()
27584
(1)
0
(2)
321052
Indirect costs ()
(3)
807494
Total budget ()
(4)
585000
(5)
IMPACT
Expected results
Airbus is one of the two leading aircraft manufacturers in the world, regularly having more than 50%
of the market share of all aircraft with more than 100 seats. Production worldwide is of the order of
40 aircraft per day. Aerospace is one of the highest value manufacturing sectors, and a critical industry
for Europe. It is also an industry that is now being to automate more and more of its core
manufacturing processes. This will be critical to maintaining Airbuss and thus Europes position at the
forefront of aircraft manufacture. Reducing lead times while ensuring quality is thus the driver for
Airbus, and it is critical for the future of the company that all routes to automation are explored and
exploited as soon as they become feasible.
The main motivation of this project is to reduce the incidence of non-ergonomic positions during
aircraft assembly, reducing work injuries, and providing workers with a better quality of life during
their work period. The obstacles are represented by difficult access areas: inside the aircraft many
work positions require movement flexibility that a SoA robot doesnt possess. In order to keep this
flexibility, the automation systems have to execute operations while sharing same physical
environment with humans. This is why there is the need for a mobile manipulation platform. Up to
now, most of the automation approaches within aircraft manufacturing have been limited to
applications outside the fuselage due to the often very complex aircraft structures and/or the confined
space.
The major exploitation route for Airbus from this project will be for the company to be able to
demonstrate a first mobile manipulation platform able to contribute to aircraft manufacture. It has
become absolutely clear during the proposal formation that the primary need for Airbus is to reduce
ergonomic stress on its workers, while increasing production speed and maintaining or improving
production quality. Airframe manufacture is unusual in that while some elements can be automated
using traditional industrial robotics the nature of airframes means that this cannot be achieved
throughout the production line. In particular many elements are still require manual dexterity, and
[confidential]
Page 21 of 26
CP-IP 608849
must be performed in cramped spaces that would be inaccessible to larger industrial robots. The
specific impacts aimed for are:
1.
2.
3.
4.
Improvement of ergonomic factors and injuries for cost reduction

Improvement of the product quality due to standard operations
Reduction of lead time
Reduction of risk and secure a continuous operation of production
Clearly it is difficult to quantify the immediate financial impact of these benefits, but a reduction in
lead time of just 1% due to extensive roboticisation of the currently fully manual parts of aircraft
manufacture would result in an additional aircraft being produced every 2.5 days with a
commensurate increase in turnover and profit. It is important to achieve this that an approach is
adopted that will lead to multiple operations being partially automated. The exploitation strategy is
to show benefit in one operation, and then expand it to other operations. This is why the transversal
seam of the aircraft has been identified as the area of application, as it requires many different
operations. In particular we aim to have impact on collaborative riveting tasks, with panel fixing as a
secondary task. Thus the approach is planned for a variety of supporting tasks (i.e., scenarios) so as to
have a multi-purpose robot that can subsequently be easily adjusted to other tasks.
Economic Benefits: It is clear that Airbuss competitors, especially Boeing, are now engaged heavily in
attempts to automate more and more of the manufacturing process. While automation in aerospace
is still relatively low, increased productivity will give huge advantages to the manufacturer staying
ahead in this technology. Airbus SAS, as one of Europes largest manufacturers, must make rapid
progress on this to ensure continued growth and market share. Since so much aircraft manufacture is
manual the impact will be best achieved with a flexible solution that extracts the maximum from the
investment. This is precisely what a multi-level autonomy approach achieves. Thus as it is adopted to
perform different tasks, a flexible mobile platform can be a central part of Airbuss manufacturing
operations. So even though some of the impacts will not all occur in the short term, the effort must
be started, and the impact on Airbuss competitiveness and market share will be vast over time. This
in turn has significant impact on the manufacturing exports of the EU.
Social Benefits: while aircraft manufacturing is a high value activity economically it involves many
manual operations that are stressful on the human worker. The benefits for the workers in terms of
working environment and occupational health are very clear. Airbus SAS are keen to achieve these
improvements, and mobile manipulation robots are a promising way to achieve this impact.
Technological Benefits: over time the level and complexity of operations performed by robots will
increase, with more of the task being performed by the robot, and less by the human. The key
technological benefit of this project is to start Airbus down this road, which will lead to a series of
robot systems performing multiple tasks. The other benefit will be in terms of improved quality of
aircraft manufacture over time.
Exploitation plan of results and management of knowledge and of intellectual

property
The exploitation plan is as follows. We will disseminate to several audiences in the project. First within
Airbus, we will disseminate the results to senior managers responsible for manufacture, and also
through the acceptance testing process with workers on the shop floor. This dissemination action will
[confidential]
Page 22 of 26
CP-IP 608849
also result in feedback to adjust the approach. We will hold a workshop to disseminate and discuss
the use case and the wider potential for mobile manipulation in aircraft manufacture at Airbus in
Hamburg.
In addition we will take, subject to the removal of Airbus specific confidential information, the
following dissemination actions to bring information about the project to a wider number of potential
users in other industries that can also benefit from joint human-robot manipulation.
1 Cross industry demonstration at MTC. Using a duplicate platform at MTC we will host a
demonstration to industrialists outside of the immediate set of competitors for Airbus. This will
involve demonstration a generic tool operation task. This will be carried out after the showcase.
2 Media coverage. UoB have an outstanding record of media engagement, and Prof Wyatt and Dr
Hawes make frequent media appearances in the local, national and international media for their work.
We will ensure that the benefits of human-robot co-workers are given maximum coverage with the
general public.
3 One-to-one Company Meetings. Through the MTC, which has more than 50 industrial members, we
will explore the exploitation possibilities in a series of one-to-one meetings with companies.
4 Airbus internal seminars. In addition to the demonstration and acceptance testing at Airbus we will
hold a small number of internal seminars with different manufacturing groups in Airbus to disseminate
the work beyond the Hamburg facility. This will be useful in exploring wider use of the approach.
5 Academic Dissemination. We will publish papers in leading international conferences and journals
on the work, while adhering to non-release of agreed confidential information.
IPR Management. Prior to commencement of the work, an IPR agreement will made with clauses to
enable smooth management of the background and foreground IP. The background will reside with
the partners, but with clauses consistent with the EuRoC appendix on IPR to enable its use in this
domain managed by the partners. The consortium will seek in the first instance to enable jointly
generated foreground IP to be applied within Airbus, without prejudicing the ability of the other
partners to exploit it in domains aside from aerospace (according to the EuRoC IPR appendix), and to
continue to exploit their background freely and independently. The technology transfer wings of the
research partners shall work with Airbus to achieve this. In addition MTC has LU and UoB as founding
research partners, and so has been established to be the natural organisation to transfer IP from low
TRLs into prototypes suitable for further funding. In Stage III of the project Airbus and the partners
will develop the joint exploitation plan covering the future development of the technology for use by
Airbus.
Ethical issues
The proposal is not believed to raise any particular ethical issues. The process for the human-robot
collaboration will undergo a careful safety evaluation prior to acceptance testing. During acceptance
testing any possible safety issues will be identified and acted on.
[confidential]
Page 23 of 26
CP-IP 608849
APPENDIX: Articles II.26, II.27, II.28 and II.29 Annex I of the EuRoC Grant
Agreement
II.26. Ownership
1. Foreground shall be the property of the beneficiary carrying out the work generating that
foreground.
2. Where several beneficiaries have jointly carried out work generating foreground and where
their respective share of the work cannot be ascertained, they shall have joint ownership of
such foreground. They shall establish an agreement1 regarding the allocation and terms of
exercising that joint ownership.
However, where no joint ownership agreement has yet been concluded, each of the joint
owners shall be entitled to grant non-exclusive licences to third parties, without any right to
sub-licence, subject to the following conditions:
a. at least 45 days prior notice must be given to the other joint owner(s); and
b. fair and reasonable compensation must be provided to the other joint owner(s).
3. If employees or other personnel working for a beneficiary are entitled to claim rights to
foreground, the beneficiary shall ensure that it is possible to exercise those rights in a manner
compatible with its obligations under this grant agreement.
II.27. Transfer
1. Where a beneficiary transfers ownership of foreground, it shall pass on its obligations
regarding that foreground to the assignee including the obligation to pass those obligations
on to any subsequent assignee.
2. Subject to its obligations concerning confidentiality such as in the framework of a merger or
an acquisition of an important part of its assets, where a beneficiary is required to pass on its
obligations to provide access rights, it shall give at least 45 days prior notice to the other
beneficiaries of the envisaged transfer, together with sufficient information concerning the
envisaged new owner of the foreground to permit the other beneficiaries to exercise their
access rights.
However, the beneficiaries may, by written agreement, agree on a different time-limit or
waive their right to prior notice in the case of transfers of ownership from one beneficiary to
a specifically identified third party.
3. Following notification in accordance with paragraph 2, any other beneficiary may object
within 30 days of the notification or within a different time-limit agreed in writing, to any
envisaged transfer of ownership on the grounds that it would adversely affect its access rights.
Where any of the other beneficiaries demonstrate that their access rights would be adversely
affected, the intended transfer shall not take place until agreement has been reached
between the beneficiaries concerned.
4. Where a beneficiary intends to transfer ownership of foreground to a third party established
in a third country not associated to the Seventh Framework Programme, the Commission may
object to such transfer of ownership of foreground, if it considers that this is not in accordance
The joint owners may of course agree not to continue with joint ownership but decide on an alternative regime (for
example, a single owner with access rights for the other beneficiaries that transferred their ownership share).
[confidential]
Page 24 of 26
CP-IP 608849
with the interests of developing the competitiveness of the European economy or is

inconsistent with ethical principles or security considerations.
In such cases, the transfer of ownership shall not take place unless the Commission is satisfied
that appropriate safeguards will be put in place and has authorised the transfer in writing.
In projects funded by Euratom, security considerations must be understood as being the
defence interests of the Member States within the meaning of Article 24 of the Treaty
establishing the European Atomic Energy Community.
II.28. Protection
1. Where foreground is capable of industrial or commercial application, its owner shall provide
for its adequate and effective protection, having due regard to its legitimate interests and the
legitimate interests, particularly the commercial interests, of the other beneficiaries.
Where a beneficiary which is not the owner of the foreground invokes its legitimate interest,
it must, in any given instance, show that it would suffer disproportionately great harm.
2. Patent applications relating to foreground, filed by or on behalf of a beneficiary must include
the following statement to indicate that said foreground was generated with the assistance of
financial support from [the Union] [Euratom]:
The work leading to this invention has received funding from the [European Union]
[European Atomic Energy Community] Seventh Framework Programme ([FP7/2007- 2013]
[FP7/2007-2011]) under grant agreement n [xxxxxx]2.
Furthermore, all patent applications relating to foreground filed shall be reported in the plan
for the use and dissemination of foreground, including sufficient details/references to enable
the Commission to trace the patent (application). Any such filing arising after the final report
must be notified to the Commission including the same details/references.
3. Where the foreground is capable of industrial or commercial application and its owner does
not protect it and does not transfer it to another beneficiary, an affiliated entity established
in a Member State or Associated country or any other third party established in a Member
State or Associated country along with the associated obligations in accordance with Article
II.27, no dissemination activities relating to that foreground may take place before the
Commission has been informed. The Commission must be informed at the latest 45 days prior
to the intended dissemination activity.
In such cases, [the Union] [Euratom] may, with the consent of the beneficiary concerned,
assume ownership of that foreground and adopt measures for its adequate and effective
protection. The beneficiary concerned may refuse consent only if it can demonstrate that its
legitimate interests would suffer disproportionately great harm.
In the event [the Union] [Euratom] assumes ownership, it shall take on the obligations
regarding the granting of access rights.
II.29. Use
1. The beneficiaries shall use the foreground which they own or ensure that it is used.
This statement will have to be translated into the language of the patent filing. Translations in all European Union
languages will be provided.
[confidential]
Page 25 of 26
CP-IP 608849
2. The beneficiaries shall report on the expected use to be made of foreground in the plan for
the use and dissemination of foreground. The information must be sufficiently detailed to
permit the Commission to carry out any related audit.
[confidential]
Page 26 of 26

Euroc Proposal MTC Lu Uob Ab

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Euroc Proposal MTC Lu Uob Ab

Uploaded by

Copyright:

Available Formats

Joint Robot-Human Logistics and

Guide for proposal preparation

Guide for proposal preparation

Challenge 2 Shop Floor Logistics and Manipulation

Challenge 3 Plant Servicing and Inspection

Challenge Team Leader (CTL)

Manufacturing Technology Centre

Manufacturing Technology Centre

Pilot Way, Ansty Park

+44 24760 701600

Wolfson School of MME

Guide for proposal preparation

Professor Jeremy L Wyatt

School of Computer Science

+44 121 414 4788

Dr. Robert Alexander Goehlich Manufacturing Engineering Project Manager

Guide for proposal preparation

Identifying and grasping required tools and materials from a work-surface.

Guide for proposal preparation

SCIENTIFIC AND TECHNICAL QUALITY

Guide for proposal preparation

Guide for proposal preparation

Guide for proposal preparation

Progress beyond the state-of-the-art

S/T methodology and associated work plan

Guide for proposal preparation

Table 1a: Task list

Round (a) of Stage II: Benchmarking+free-style

Round (b) of Stage II: Showcase round

Stage III (Field tests Pilot Experiments)

Tables 1b: Brief description of tasks

Round (a) of Stage II: benchmarking+free-style (1)

Description of work and role of participants

Guide for proposal preparation

Round (b) of Stage II: showcase round (1)

Description of work and role of participants

Guide for proposal preparation

Stage III (Field tests Pilot Experiments) (1)

Description of work and role of participants

Guide for proposal preparation

Table 1c: List of deliverables

Task(s) Nature Dissemination

Report describing the activities of the team and

Report describing the activities of the team and the

Report describing the activities of the team and the

Table 1e: List of milestones

Report describing the activities of the team and

Guide for proposal preparation

Round (b) of Stage II

Report describing the activities of the team and the

Round III completed

Report describing the activities of the team and the

Table 1f: Summary of staff effort

Task3 Total person months

Table 1g: Risk assessment and contingency plan

Force assistance is unnatural

ORevert to gravity compensate or only apply

Difficult access areas also

Use only robot for areas that it can reach to

Robot arm cannot reach

Use only positions that are in reach with robot arm

Shdow hand cant be used

We have access to a lightweight Pisa-IIT hand that