Cloud-Based Automated Design and Additive Manufacturing: A Usage Data-Enabled Paradigm Shift

Additive and adaptive Manufacturing with back propagation of sensing data using mobile Additive and adaptive Manufacturing with back propagation of sensing data using agents from robots to the design and iteration process resulting in continuous series improvements. mobile agents from robots to the design and iteration process resulting in continuous series improvements.?

Abstract:

Integration of sensors into various kinds of products and machines provides access to

in-depth usage information as basis for product optimization. Presently, this large potential for

more user-friendly and ef?cient products is not being realized because (a) sensor integration and

thus usage information is not available on a large scale and (b) product optimization requires

considerable efforts in terms of manpower and adaptation of production equipment. However,

with the advent of cloud-based services and highly ?exible additive manufacturing techniques,

these obstacles are currently crumbling away at rapid pace. The present study explores the state

of the art in gathering and evaluating product usage and life cycle data, additive manufacturing

and sensor integration, automated design and cloud-based services in manufacturing. By joining

and extrapolating development trends in these areas, it delimits the foundations of a manufacturing

concept that will allow continuous and economically viable product optimization on a general, user

group or individual user level. This projection is checked against three different application scenarios,

each of which stresses different aspects of the underlying holistic concept. The following discussion

identi?es critical issues and research needs by adopting the relevant stakeholder perspectives.

Keywords:

sensor integration; PEID; PLM; additive manufacturing; cloud-based manufacturing;

engineering design; product customization; product design; product development; automated design

1. Introduction

Imagine there’s a product that starts collecting data from the moment that it’s being made—a

product that collects, then evaluates these data, and passes the results on into the cloud: information

Sensors 2015,15, 32079–32122; doi:10.3390/s151229905 www.mdpi.com/journal/sensors

Sensors 2015,15, 32079–32122

that describes its making and its everyday experiences. In the cloud, this evidence is matched with

what other individual products of identical type and make provide, and scrutinized on an item and a

type level. In the cloud, this information is processed to automatically adapt design and dimensioning

of coming product generations. But why wait until a coming generation? And, why, speci?cally, do so

if all the information needed is readily at hand to optimize the product not only for some general set of

requirements, as sophisticated as they may be, but for an individual customer, and an individual use

pattern? Why not allow the product to evolve, not from generation to generation, but from item to

item? And do so differently from individual customer to individual customer, too?

Figure 1graphically represents the impact of this concept on product generations and diversity,

with continuous change replacing the major revisions (these are retained in the new approach on

lower level with e.g., feature addition as discriminator, leading to product groups with close internal

and more remote external relation) as well as the minor ones of a more common organization of the

process. Conventional production technology, speci?cally in mass production, is not matched easily

with this vision. Mass production typically relies on complex manufacturing equipment and thus high

investment costs to reduce part costs based on an economy of scales approach: Flexibility is sacri?ced

for the sake of productivity. Tools are matched to products, causing adaptation of the latter to be costly.

What, then, if production processes provided boundless ?exibility, and changes to the product could

be realized virtually, as changes to its digital representation, and at virtually no cost?

Sensors 2015, 15, page–page

2

1. Introduction

Imagine there’s a product that starts collecting data from the moment that it’s being made—a

product that collects, then evaluates these data, and passes the results on into the cloud: information

that describes its making and its everyday experiences. In the cloud, this evidence is matched with

what other individual products of identical type and make provide, and scrutinized on an item and

a type level. In the cloud, this information is processed to automatically adapt design and

dimensioning of coming product generations. But why wait until a coming generation? And, why,

specifically, do so if all the information needed is readily at hand to optimize the product not only

for some general set of requirements, as sophisticated as they may be, but for an individual

customer, and an individual use pattern? Why not allow the product to evolve, not from generation

to generation, but from item to item? And do so differently from individual customer to individual

customer, too?

Figure 1 graphically represents the impact of this concept on product generations and diversity,

with continuous change replacing the major revisions (these are retained in the new approach on

lower level with e.g., feature addition as discriminator, leading to product groups with close internal

and more remote external relation) as well as the minor ones of a more common organization of the

process. Conventional production technology, specifically in mass production, is not matched easily

with this vision. Mass production typically relies on complex manufacturing equipment and thus

high investment costs to reduce part costs based on an economy of scales approach: Flexibility is

sacrificed for the sake of productivity. Tools are matched to products, causing adaptation of the

latter to be costly. What, then, if production processes provided boundless flexibility, and changes to

the product could be realized virtually, as changes to its digital representation, and at virtually no

cost?

Figure 1. Consequences of the basic concept visualized: From product generations to continuous

optimization through gathering and usage of life cycle data in conjunction with flexible production.

In a manufacturing environment of this kind, the making of the product would not be linked to

the physical site at which dedicated tools and machinery were kept, simply because no such tools

and machinery would be needed. Instead, availability of the digital product information alone

would enable countless manufacturing centers worldwide to create the product without lead-time.

This is why such approaches have been termed Direct Digital Manufacturing [1].

A global manufacturing environment like this could transform the way we make things. The

paradigm shifts implied are manifold: For one thing, as product development would not require

parallel development of production equipment, design and production could move further apart. At

Figure 1.

Consequences of the basic concept visualized: From product generations to continuous

optimization through gathering and usage of life cycle data in conjunction with ?exible production.

In a manufacturing environment of this kind, the making of the product would not be linked

to the physical site at which dedicated tools and machinery were kept, simply because no such tools

and machinery would be needed. Instead, availability of the digital product information alone would

enable countless manufacturing centers worldwide to create the product without lead-time. This is

why such approaches have been termed Direct Digital Manufacturing [1].

A global manufacturing environment like this could transform the way we make things.

The paradigm shifts implied are manifold: For one thing, as product development would not require

parallel development of production equipment, design and production could move further apart.

At the same time, and as described above, ?exibility in manufacturing could be used to optimize

products on a customer or customer group basis, and furthermore, to implement a continuous design

optimization. Practically this means nothing less than brushing aside the fundamental concept of

individual product generations. Since optimization needs a basis, a primary prerequisite is a usage

32080

Sensors 2015,15, 32079–32122

data link allowing back?ow of information from product to designer. Equally important is a legally

and technically secure access for the various potential producers and service providers to initial and

optimized product design and manufacturing information. This global access to (general product and

usage) data, software tools and ?nally manufacturing resources is the primary link of our scenario

to the concept of the cloud (please consider also Figure 8in Section 4in this respect). Cloud-based

manufacturing (CBM), sometimes also designated Cloud Manufacturing (CMfg), in general is currently

the object of intense study [

2

–

5

] Xu et al., to give an example, introduced cloud manufacturing via

the earlier-adopted concept of cloud computing, de?ned by the National Institute of Standards and

Technology (NIST) as “a model for enabling ubiquitous, convenient, on-demand network access to

a shared pool of con?gurable computing resources (e.g., networks, servers, storage, applications,

and services) that can be rapidly provisioned and released with minimal management effort or

service provider interaction.” [

6

]. Cloud manufacturing extends Cloud Computing by including

production processes (scheduling, resource planning etc.) and related Cyber-Physical Systems (CPS)

as active units [

3

]. Another de?nition has been provided by Wu et al., who discuss whether CBM

is indeed the paradigm change as which it is currently being advertised. Their answer is in the

af?rmative: By reviewing the suggested de?nitions of the ?eld and adding their own perspective, they

manage to delimit CBM both from earlier concepts like ?exible and redistributable manufacturing

systems and intermediate development stages like web- and agent-based manufacturing (WBM,

ABM) [

7

]. A true CBM approach, in their eyes, needs to integrate “Infrastructure-as-a-Service

(IaaS), Platform-as-a-Service (PaaS), Hardware-as-a-Service (HaaS), and Software-as-a-Service (SaaS)”

elements and is distinguished from WBM and ABM by its capability of facilitating new business

models through such elements [

7

,

8

]. In further publications, Wu et al. have linked their studies to

additive manufacturing and looked at product design, too, thus extending the original term CBM to

cloud-based design and manufacturing (CBDM) [

9

,

10

]. Interestingly, however, the usage data feedback

is not considered a core feature of CBM in these de?nitions: Data from the usage or Middle-of-Life

(MoL) phase is included in some of the projections offered, though not in an automated fashion, but

solely by way of direct end user (customer) feedback and integration in the design process (customer

co-design). The practical implementation of such end user-centered feedback facilities is seen as major

research issue. Besides, as we will show later in case study 2 (Section 4.2), the notion of separated ABM

and CBM approaches brought forward by Wu et al. can be overcome [9,10].

All the aforementioned capabilities require sophisticated analysis of data as glue between the

various enablers on technological level. On a more generic level, Arti?cial Intelligence (AI), Machine

Learning (ML) and advanced ICT are already integrated in most areas of daily life, as well as in

industrial production and product development. In parallel, the value of information and data is

rapidly increasing—more so if they help in satisfying customer needs. Companies who understand

the needs of their customers and at the same time are capable of translating them in a timely

manner into new products and product enhancements, have a competitive advantage in the global

business environment.

This statement accepted, several equally important aspects are required to succeed in the future:

?Availability of information about the usage of individual product items.

?Capability to analyze and translate this information into technical requirements.

?Capability to map these requirements to product design (e.g., CAD model).

?Capability to economically manufacture the products designed accordingly.

?Capability to increasingly automate these steps to realize fast time-to-market.

That said, we may match our own perspective of what might be called “Usage Data-Enhanced

Cloud-based Design and Manufacturing” or UDE-CBDM to the requirements proposed in the literature

to distinguish cloud-based from conventional manufacturing. We have done so in Figure 2, contrasting

the set of eight requirements a CBDM approach has to meet according to Wu et al. [

8

] with two

additional requirements that together with the initial ones constitute the UDE-CBDM case.

#snsinstitutions

#snsdesignthinkers

#designthinking