IEEE Systems, Man, and Cybernetics Magazine - January 2018 - 13

behind the scenes to acquire and infer context information.
production processes can be redesigned to achieve high
Some wearable or embedded sensors may be used to get
efficiency. With real-time information collected, intelligent
users' exact requirements, e.g., to detect their health condialgorithms, and networked actuators, the control software
tions and predict their behaviors. Such autonomous intercan automatically make decisions and drive actuators to
action may be continuous so as to adapt the working
shrink the deviations from the plan. Large amounts of mulenvironment according to the changing user needs.
tisource data and intelligent machine-learning algorithms
Three-dimensional reconstruction and interaction reprecan automatically generate optimal decisions. The advance
sent a future trend that can provide vivid and immersive
of machine-learning technology substantially increases the
experience [46]. In an idealistic scenario, factory workers
level of autonomy to control production processes and deal
can talk with reconstructed images of their managers anywith various disruptions.
where and anytime, and it would be just like they are talking
in the same physical location. Emotional factors are also
Energy Management and Green Manufacturing
important for human-machine interactions and virtual/augManufacturing accounts for about one-third of global enermented-reality-based remote human-human interactions
gy demand [50]. When coupled with increasing energy pric[47]. In the long term, sense and emotion will be combined
es, energy management is not a trivial issue. Traditional
to construct an advanced virtual
methods are based on isolated
collaboration environment, where
plant states without a full underusers can feel that humans and/or
standing of the whole plant due to a
Latency-sensitive
machines work at the same physical
lack of infrastructure for holistic
site. The humans and machines of
mapping to business and fineapplications require
interest are pervasively presented
grained, continuous measurement
the collaboration of
around users.
of energy consumption. Not only
can the IoT help to continuously
end devices (e.g.,
Other Critical Issues
track and correlate energy conWSNs, mobile phones),
In manufacturing, there is a large
sumption and business activities in
number of latency-sensitive applireal time by deploying sensors at
mediate nodes (e.g.,
cations that request real-time perany locations of interest; it also
the base station in
ception, decision making, and
enforces online dynamic energyWSNs, gateways),
actuation. This requires the collabaware control in the IoT-enabled
oration of end devices (e.g., WSNs,
closed loops.
and data centers.
mobile phones), mediate nodes
Energy efficiency should go
(e.g., the base station in WSNs, gatebeyond simple stand-alone ap ways), and data centers. For the
proaches, e.g., single process/
(powerful) mediate nodes, the paradigm has a name, fog
machine optimization, toward a more holistic view.
computing or edge computing [48], that complements and
Cross-domain collaboration (in the physical world, e.g.,
extends the cloud computing paradigm to the edge of the
machinery, materials, and vehicles, and in the business
network, with characteristics such as low latency, location
world, e.g., enterprise information systems, production
awareness, and strong presence of streaming and real-time
processes, and logistics) and data acquisition and correapplications. It uses field-area networks at the edge to facillation must be in place to develop good strategies. Also,
itate the machine-to-machine or human-to-machine interstatistical analysis and real-time energy-related indexes
actions. Additionally, it filters data to be consumed locally
should be combined as a whole. Big data analytics can
and sends the rest to the higher tiers.
play an important role in moving in the direction of
Privacy and security issues are crucial in a future open
green manufacturing.
and highly connective world. We have conducted a comprehensive literature review, which is presented in [49].
Proactive Maintenance
Manufacturers have widely accepted the concept of proacFuture Applications of the IoT in Manufacturing
tive maintenance, which advocates early diagnostics and
part replacement based on the prediction and monitoring
Automation and Production Efficiency
of machine degradation to reduce costly, unscheduled
IoT systems collect real-time status data from the factory
downtime and unexpected breakdowns [51]. Lower-cost
floors (e.g., machinery, vehicles, materials, people, and
sensors, wireless connectivity, and big data tools can
environments) and feed them into enterprise decision-makdeliver useful data and analysis about a machine's status
ing systems. Those data can be used to automate workand performance. Historical and real-time data can be
flows/processes to maintain and optimize design and
modeled, correlated, analyzed, and visualized to make
production systems without human intervention. For
machine degradation predictable and visible. Also, such
example, through process mining powered by the IoT,
data can be fed back to product designers for closed-loop
Ja nua r y 2 01 8

IEEE SYSTEMS, MAN, & CYBERNETICS MAGAZINE

Table of Contents for the Digital Edition of IEEE Systems, Man, and Cybernetics Magazine - January 2018