#datalift use cases in production in mobility
Let's break the proof-of-concept cycle and productionize data analytics and machine learning.
With #datalift, we show you how business and government can scale the data economy now.
In this post, we highlight use cases in production presented by AI Guild members who are experts in deploying in the mobility space.
ML-based object recognition for autonomous driving
Presented by Julia Nitsch, AI System Architect, Ibeo Automotive
Moderated by Andrés Prada González, Computer Vision Engineer, Footprint Technologies GmbH
Julia Nitsch studied information and computer engineering at TU Graz, Austria. She received her M.Sc. in 2016 with a major in autonomous robotics. After her studies, she worked in the Robotics and Perception Group, University of Zurich on urban search and rescue robotics. Since 2016 she is employed at Ibeo Automotive Systems GmbH and did her Ph.D. studies in cooperation with the Autonomous Systems Lab, ETH Zurich.
Autonomous vehicles require a robust perception and reliable object recognition to ensure safe driving maneuvers. Therefore, the vehicles are equipped with a multi-modal sensor setup that provides a rich representation of the surrounding environment and ensures a robust environmental perception. Furthermore, semantic information must be extracted from the raw data to provide information about objects surrounding the ego vehicle. Data-driven concepts like Machine Learning based algorithms achieve state-of-the-art results in object recognition. However, these algorithms require 1000s of hours of recorded data for development and qualification, which creates new challenges. First, 1000s of hours of data are likely to acquire Petabytes of data which derives new requirements on the infrastructure. Secondly, required scenarios and a variety of road users must be represented in the collected data sets. This requires semantic information from raw data which can be achieved by manual labeling or algorithms processing the data.
We talk about data infrastructure and how we deal with this immense amount of data within this presentation. Due to the recording of raw data we face different challenges compared to other use cases and elaborate on them. Furthermore, we speak about the data life cycle and how data-driven algorithms can be developed in such a setup.
Watch the recorded live session here:
Digital multi-market lead generation model in automotive
Presented by Anne Schmucker, Data & Search Specialist at Mercedes-Benz AG
Moderated by Johanna Viktor, Senior AI Specialist, Deutsche Bahn
Anne Schmucker is a Data Strategist & Digital Marketing Specialist at Mercedes-Benz AG, leading the Region Europe Data Strategy team for 15+ markets. She has been working in the automotive industry for more than six years, having had diverse international project assignments. She is currently completing her Research Masters' in AI and Data Protection Compliance at Stuttgart Media University.
Intelligent online customer acquisition will define the future competitiveness of global companies across major consumer products industries, including the automotive industry. Lead generation models have already demonstrated their importance for digital marketing strategies, forming a well-established key pillar for companies’ prospect acquisition. A homogeneous, multi-level and cross-market strategy approach is key for a successful international deployment and an ongoing challenge for global players. The present keynote showcases a conceptual approach to how to deploy a lead generation model across multiple markets by harmonizing different levels of competencies and skills from different stakeholder groups and departments.
Watch the recorded live session here:
Railway maintenance and autonomous vehicles: more similar and more different than you’d think
Presented by Irina Vidal Migallón Technical Lead AI & Computer Vision, Siemens Mobility
Moderated by Sven Krüger, Co-founder & CCO, whoelse.ai
Irina Vidal Migallón is one of Europe’s leading Computer Vision practitioners. She compares the use cases of railway maintenance and autonomous vehicles, which seem similar from a technical point of view but have very different deployment requirements. Irina shows us how machine learning techniques are embedded in the use case and that the product components from the sensors to legal regulations have a profound impact.
Monitoring the state of rail tracks and autonomous driving have many tasks in common: detect obstacles, detect people, detect assets, detect damages. As pre-production work, it may seem a no-brainer to approach them in similar ways. The devil is in the details, though, which happens to be 90% of the product. Sensors, deployment platform, processing hardware, feedback loop, laws: they have a more significant impact than the machine learning technique itself.
Siemens Mobility has a division focusing on computer vision and machine learning for different products around mobility, especially also in the railway domain. Products range from passenger assistance to autonomous driving. For example, we increase comfort for travelers by finding spots for a wheelchair or detecting aggression to improve safety. For driving, our approach includes handling typical issues like obstacle detection and track segmentation.
Read the full use case in the #datalift ebook in Kindle.
Make #datalift Season 2 happen!
Across more than ten industries, from leading corporates to leading startups, and on to new firms, we have heard from those having use cases in production. With a loyal audience of practitioners, of which more than half have roles as Seniors, Leads, Directors, and CxO, the #datalift event series from No 1 to 5 has become the key forum for exchange on what works and what is best practice.
#datalift is the online meeting space where the audience most interested in closing the gap to deployment gathers.
The AI Guild would like to have your company as a partner in growing the ecosystem of deployment: we offer year-round visibility for your content, presence in live events, qualified leads, and analytics. You achieve recognition with professionals across Europe for your solution, team, and data strategy by supporting the drive for more deployment.
The partnership is for the whole season or for as many events as your company likes. Partnership starts from €1000 per event, and the range is to €63000 for the premium season ticket.