Programming artificial intelligence for object recognition made easy – by the BMW Group

Munich. The BMW Group is releasing another software
package that makes light work of creating artificial intelligence (AI)
applications for object recognition. The core of the latest release
(github.com/BMW-InnovationLab[1])
is the so-called BMW Labeling Tool Lite. This tool allows for users to
easily label objects in photos, an offline solution used in production
to quickly create AI apps that reliably identify objects in photos.

The new release is influenced by important practical experience from
the BMW Group’s production network: “Smart AI solutions make the
day-to-day work of our associates noticeably easier. Users need no
longer worry about the underlying technology. The logic is the same as
with a good smartphone app: easy to install, quick to understand, use
it of your own accord. That’s the only way a solution will quickly
attract widespread use and become more effective,” says Michele
Melchiorre, Head of Production System, Planning.

Programming artificial intelligence for object recognition made easy – by the BMW Group.

Building an object-recognition app, even with no AI expertise, can be
quick and straightforward to do, without having to programme any
software. To train an app that suits their needs, production
associates start with taking and labelling
photos. The software then optimises itself independently and can
distinguish between “right” and “wrong” after just a few hours, having
worked through the labels. By comparing live images from Production,
the app can recognise quickly and reliably whether the right parts
have been used, for instance: as proven by an AI app for accurate
recognition of up to 10 different BMW 3 Series Sedan door sill strips
at BMW Group Plant Munich.

An associate labels photos from door sill strips of the BMW 3 series, creating an AI object-recognition app.(09/2020)

“As well as supporting quality work in Production, this self-service
for AI apps also offers particular benefits for the numerous Smart
Transport Robots in Logistics at the BMW Group,” emphasises Dirk
Dreher, Head of Logistics Planning.  

Jimmy Nassif, Head of IT Planning Systems in Logistics, agrees: “With
our software package, it takes just a few hours to build apps for
comparing actual and target statuses.”  Matthias Schindler, Cluster
Supervisor for Smart Data Analytics in the production system, adds:
“And there are countless possible uses for these apps.”

The published algorithms are freely available for software developers
worldwide to use, view, modify and develop the source code further.
These developments will also benefit the BMW Group. What’s notable
about this now freely accessible software package is its simple,
straightforward use based on plug & play. Users require neither
programming skills, specific hardware or additional software; a
standard powerful PC is enough.

The BMW Group already published selected algorithms from this area of
AI back in autumn 2019. “The wealth of feedback on the algorithms we
released in 2019 was overwhelming. Our BMW AI community is delighted
with the appreciation we got from around the world. We are seeing
useful enhancements based on our source code. That prompted us to
publish more algorithms, to help open up AI for mainstream users,”
says Kai Demtroeder, Head of Data Transformation, Artificial
Intelligence in BMW Group IT.

With this latest release, the BMW Group is offering a complete
solution for AI-based object recognition. Users who value high system
stability will appreciate the additional functions such as failover
and load balancing, which have now been added to the
object-recognition interfaces (API).

The BMW Group employs a whole host of AI apps in Production and
Logistics, as they make life easier for associates by taking over
especially monotonous or tiring monitoring tasks.

References

  1. ^ github.com/BMW-InnovationLab (github.com)
1 2 3

Share