12 OPINION
Eifrem
Graph technology is invaluable for building a picture of the inter-relationships
between car components, from traction control, to power-steering and braking,
which provides engineers with fast access to critical insight, says Emil Eifrem
›From aerodynamic electric vehicles to autonomous
cars, automotive design is continuously evolving
and becoming increasingly complex as it depends
increasingly on software and connectivity to get from
A to B. In today’s family car there are around 2,000
software components and applications that process and
exchange around 10,000 signals from all parts of the
vehicle. Semi-autonomous and driverless vehicles will
go several steps further, with sophisticated systems
incorporating high-performance computers, ADAS
sensors, LIDAR and HMI.
Keeping an accurate trail of all these interrelationships
become increasingly difficult, with increased complexity to
contend with. Currently manufacturers utilise spreadsheets
and bills of material (BOM) to track them. They also keep
structured and unstructured data in enterprise resource
planning (ERP), Power Distribution Modules (PDM),
digital asset management (DAM) and product information
management (PIM) solutions. This data is then modelled
using different development tools.
While this information builds a picture of a single product,
its actual genetic history is often sealed off in data silos.
This means it is impossible to quickly and easily access
all the data required to see how individual automotive
components are connected. This challenge is getting
bigger because it is these same connections that identify
a component’s role in the vehicle, points of failure
and reliance.
The magnitude and complexity of component data is
becoming a real issue. This data needs to be readily available
and actionable. Graph database technology has the answer
to this conundrum in its power to record complex data
On a ight to Mumbai
in 2000, Emil Eifrem
sketched what today is
known as the property
graph model. As
CEO and co-founder of
Neo4j, he has devoted
his professional life
to building and
evangelising graph
databases. Committed
to sustainable open
source, Emil plans to
‘save the world with
graphs’ and own
Larry Ellison’s yacht
VehicleDynamicsInternational.com • November/December 2019
interdependencies that generates insight for smart decision
making. This is because it uses notational formalism closely
aligned with the way humans comprehend information.
If you are unfamiliar with this approach, a graph database
is a way of managing data that is unlike the traditional
relational databases from vendors such as Oracle and Microsoft.
Graph technology is a powerful way to look under the
covers of complex systems. Once the data model is coded in
a scalable architecture, a graph database is almost impossible
to beat at analysing the complex relationships between a large
number of data points.
This approach allows automotive engineers to better manage
and visualise large data sets and extract value from them –
mapping product components and their inter-relationships to
evaluate risk of failure, for example, while providing consistent
and accurate trackability.
Embedded in standard product data management (PDM),
a graph database can create a map encompassing all the
functions of a product and then graphically trace the complex
data interdependencies. Instead of getting a ‘laundry list’ of
all the product’s functions, a central knowledge hub is built to
map the total product. Using analytical methodology, clusters
and loops in the system design can also be swiftly identified.
This data explosion is changing the landscape for
automotive manufacturers. The automotive industry is in a
race to build smarter, more intelligent cars, which is driving
complexity. At the same time, car makers need to become more
agile in their design processes as competition becomes more
intense. Graph technology finally gives automotive engineers
and designers a way of managing these large, highly complex
data sets to improve functionality and come up with
innovative new features.
“It is impossible to quickly and easily access all the data
required to see how individual components are connected”
/VehicleDynamicsInternational.com