INTELLIGENCE
it is involved. Such data may
be gleaned from a variety of
sources, ranging from sales and
booking enquiries to physical
sensors and Internet of Things ARTIFICIAL
Freight sector gAIns
Air cargo is facing the AI revolution – but is it prepared? Adrian Kosowski, who is the
Head of Research at NavAlgo, asks the big question.
Innovation comes in waves.
The forerunners of change,
after a decade or two, tend
to lag behind. The air cargo
industry, with a business process
essentially unchanged since
the 1980s and with close to
half of all air freight still being
processed through paper air
waybills, can today only be
described as a very traditional
industry, one that is waiting for
a wave of innovation.
Air cargo is a traditional
industry, operating in a
rapidly changing economic
environment. For one thing,
it has a rather complicated
relationship with actors of the
New Economy. E-commerce
giants, such as Amazon
and Alibaba, account for an
increasing proportion of carried
cargo, but are also not shy of
aligning the entire logistics
process around their operations.
Today, Amazon’s Prime Air
runs physical operations which
are in direct competition with
air freight carriers and freight
forwarders alike, while Alibaba’s
intelligent logistics arm,
Cainiao, spins a vision of global
transportation, potentially
reducing the role of carriers to
performing precisely planned,
subcontracted operations. In
terms of scale, this is not an
unrealistic ambition: on one
special day this November,
Alibaba topped US$40bn in
sales, more than the entire air
cargo industry made this year in an average calendar quarter.
At the same time, declares Kosowski, the air cargo industry is
being challenged by other modes of transportation. New shipping
services with intelligent containers, sea-based or intermodal, and
increasingly featuring a rail segment between Asia and Europe, offer
a level of cargo visibility and traceability which is winning over the
hearts of customers – and drawing them away from air freight.
Adapting to change
It is clear that the air cargo industry needs to adapt to these changes
– but how? As any much-hyped keyword, Artifi cial Intelligence is
entering the air industry through fl ashy but isolated use cases. A
recent report by SITA concluded that by 2021, a vast majority (85%)
of passenger airlines plan to use AI to improve communication with
their clients: for example, by adding chatbots on an airline website.
At the same time, a full third of airlines have not envisaged even a
single use case concerning predictive analytics, and those that do
envisage such use cases, typically focus on predictive maintenance
of equipment. This is all very far from a global vision of intelligently
planned logistics operations driven by data, which is being drawn
up by actors of the New Economy. A vision, where cargo fl ow
forecasts obtained using state-of-the-art machine learning methods
form the basis for optimising operations, for all actors, from airlines
to ground crews. A vision, in which the main value delivered to the
client concerns readily accessible information, involving traceable
cargo, delivered reliably and on time. A vision which will dominate
world trade in the 2020s – whether the current market leaders decide
to participate in it or not.
Dangers to the cargo sector
Today, it is the experts of the commodity market – the cargo actors
themselves – who, rather ironically, run the risk of commoditisation.
Without clear differentiating factors in terms of services offered
to clients, a logistics service could all too easily be reduced to a
small number of factors: typically price and time. Such factors
have naturally contributed to the rapid growth of third party price
comparison and booking platforms, such as cargo.one, highlighting
the increasingly fragile nature of customer relationships in the
industry. Unlike in passenger transport, the transported cargo does
not show much gratitude towards its caregivers, and there are no
welcome drinks or lounges in cargo terminals, which would help to
build client attachment.
Lessons to learn
Based on our experience across
different modes of transport,
we fi nd that there are several
lessons which appear to hold
universally.
First, it always pays to “act
globally”, and in particular, to
favour global data processing
on an entire logistics network
over planning operations in
a pointwise way. The second
ingredient is to always rely on
your own data, while working
with other actors.
This is not to say that
current standardisation efforts
for air cargo or industry cooperation
in terms of data
collection and sharing are not
crucial for the sector – they
are, especially in areas related
to chain of trust and shipment
insurance. Nevertheless,
collaborative efforts may
come too late, and the stake
in the game is too high to
lose through inertia in the
sector. It is very clear that an
airline should not, in general,
place an attempt to integrate
data from multiple airport
ecosystems at the heart of
its AI strategy, giving instead
higher priority to data which
refl ects the nature of the endto
end supply chains in which
42 December 2019 www.airlogisticsinternational.com
© BAIVECTOR - stock.adobe.com
/www.airlogisticsinternational.com
/stock.adobe.com