Mobility and the Road Ahead - S9

of computation, automation and communication onto devices that
are mobile, " he said.
Aaron Jefferson, vice president of product at Luminar Tech, suggests
an even simpler model: " Mobility is, quite simply, how you get
from point A to point B. " But how you get there today requires more
technology-dependent systems that include automation, autonomous
activity, efficiency, green technologies and safety measures to
match these advancements.
Automation efficiencies in the real world
Automation is vital to the future of advanced mobility, but differences
exist in how automation is used in the practical world, such as in urban
versus rural environments and in first- versus last-mile delivery.
" Last mile is the most expensive leg of the entire delivery journey, and
minimizing the variable costs upfront opens up delivery options that
would otherwise be very prohibitive for a wide range of merchants, "
Patel said. " E-commerce is slated to grow
by multiples in this decade because of [the]
flexibility of delivery options. ... [These]
become economically feasible options for
logistics enabled with automation. "
But can there be too much of a good thing
with automation? With the technology
must come new security requirements,
including vehicle system safety, critical
operations and standards, which require
mobility players to rethink systems.
Peter Brown, chief automotive architect
of software and controls at Ricardo North
America, points out that mobility is often focused on helping individuals
get where they want to go and deliver the products they need.
Even so, he suggests that greater security precautions must be
taken along the way. " It still concerns me that cybersecurity is not
where it needs to be to reach the levels of autonomy that we want. "
the requirements increase to implement and enable autonomy, so,
too, will the technologies that enable the computing solutions to
continue to shrink. From a platform software perspective, this points
to exciting times ahead.
" Virtually all the OEMs are doing wholesale redesigns of their
architecture, " said Artur Siedel, vice president of the Americas at
Elektrobit. " We actually really see the next wave of platforms in the
2023-24 time frame. And that's where you will really see ... software
stacks that basically are consistent and updatable. "
" We want to enable our
customers to innovate. We
want to enable functions
and capabilities that the
user would want to pay
for and don't want to live
without. "
- Artur Siedel, Elektrobit
What's driving automation?
Electric-vehicle technology is one of many enablers for the widespread
advancement of automation to full autonomy, with greater
compute power for connectivity and data processing.
Brown suggests that driving down costs and driving up reusability
are the first steps to true autonomy. " Reusability of software in
automotive has often been a particularly difficult
pain point, " he said. " I need not only to be able to
reuse software over time, but I need to be able to
have that software be well-defined and be able
to be interoperable with other systems, so standards,
and in my mind, service-oriented architectures,
are going to be key to enabling all of this. "
Several foundational architectural approaches
are being applied to scale autonomous vehicle
development, from R&D initiatives to production.
" The big trend right now is ... you see a lot
of OEMs hiring software developers, building
in-house software skills, " Siedel said. A trend
which must be coupled with collaboration and standardized development
to achieve success.
Impacts of software and automation
Jefferson said there's an environmental upside to new software and
automation in mobility, especially in electrification and AI. " If cars
become part of a fleet and you really utilize the technology as efficiently
as possible, " he said, " you really gain the efficiency of less
traffic potentially, fewer accidents, less emissions, less waste. "
Automated vehicles today still require significant onboard computer
power. However, Brown thinks that the next game-changer is starting
to take shape.
" With increased connectivity, such as 5G, maybe the solution is to
try to take some of the compute out of the vehicle itself, " he said.
" So, maybe a drone doesn't have all of the Artificial Intelligence built
onboard. It doesn't do all the sensing and sensor fusion. Maybe it
can relate or respond and be controlled remotely to try to reduce the
space, weight and power of the device. "
Such advancements in automation as a function of mobility are
expected to affect vehicle architecture. Jefferson believes that as
Setting software and autonomy
standards
In urban mobility, infrastructure and cross-vertical requirements
are necessary to make production scale for autonomous driving
a reality. States are starting to adopt different rules, for example,
which makes adoption tricky.
To effectively establish standards, Jefferson suggests that a structured
organization must be established with capabilities to keep
the industry moving in the right direction, including monitoring
functions, an ability to do forensic analysis, intrusion detection and
a guardrail to help OEMs make sound investment decisions.
The road ahead: Collaboration and
consolidation
Production-scale deployment of fully autonomous vehicles requires a
standardized software platform to enable OEMs to focus on differentiated
development. Development of the physical hardware and systems
on the vehicle side must advance concurrently with the algorithms
to help standardize the production deployment for autonomous
vehicles across the industry.
MOBILITY AND THE ROAD AHEAD *
9
Mobility and the Road Ahead

Table of Contents for the Digital Edition of Mobility and the Road Ahead
