Pilot project of Deutsche Bahn and brighter AI: Anonymized passenger analysis to comply with corona distance regulations

The implementation of distance regulations to contain COVID-19 is particularly difficult in the public transportation sector.

In order to prevent trains from overcrowding, plan passenger capacities optimally and manage passenger flows intelligently, Deutsche Bahn will be piloting automated passenger density analysis for the first time from 27 April 2020 in Stuttgart. The anonymization technology of brighter AI makes it possible to use the existing camera infrastructure of the trains in a data protection compliant way.

Strategies and scenarios for relaxing the Corona behaviour rules for a controlled ramp-up of public life and economic activity after the shutdown are currently being discussed throughout Germany. Public transport plays a central role as a system-relevant infrastructure and important public space – the use of technology can make a significant contribution to contain the wave of infection and to comply with relevant public guidelines. Dr. Dirk Rothenstein, CEO of S-Bahn Stuttgart, the executing entity, explains: “We are trying to find out whether we can use the existing technology and thus create added value for our passengers”. In a concrete application, brighter AI provides DB with the “Identity Protection Suite”, including anonymization, analysis and conformity check.

Identity Protection Suite (Edge)

In the first step, the data from existing train cameras will be made usable by means of ‘Deep Natural Anonymization’, in which personal information such as faces will be replaced by synthetic objects. The result is video recordings that protect the personal identity and at the same time enable digital processing of these datasets. An analysis is then carried out solely on the basis of protected identities and no video data ever leaves the devices. In addition, the outgoing metadata are subject to a conformity check, which ensures that no personal reference can be made (e.g. in accordance with ‘k-anonymity’).

With our anonymization technologies we want to resolve the alleged ethical conflict between privacy rights on the one hand and data-based innovation on the other. By protecting personal data, such as faces in image and video data and then making it available for analytics, we enable numerous use cases and new business models under the protection of personal identity.

Marian Gläser, co-founder and CEO of brighter AI

brighter AI is a portfolio company of eCAPITAL AG. eCAPITAL has been the lead investor and together with G+D Ventures invested in the company’s seed financing round in July 2019. “Any processing, including the transmission and storage, of video data in our modern society must meet data protection requirements. The team at brighter AI succeeds time and again in anonymizing video data in a particularly precise, reliable and efficient manner,” says Dirk Seewald, Partner at eCAPITAL. “This technology Made in Germany is a central component of data protection-compliant video and image data processing.”

Deep Natural Anonymization vs Traditional Blurring

Pilot projects along the lines of the passenger analysis of DB and brighter AI make an indispensable contribution to establishing new standards and creating social acceptance of data-based services and products. For example, the cooperation between the start-up and the long-established company is subject to a continuous data protection review of anonymization and the downstream analytics purposes. The parties are convinced that where the use of technology provides visible social added value and personal rights are simultaneously protected to a high degree, sustainable models are created and the way is paved for a digitization practice that does not play off data protection and innovation against each other.

About brighter AI
brighter AI Technologies is a Berlin-based startup founded in 2017. brighter AI is working on to overcome limitations in human and computer vision with artificial intelligence. With an innovative anonymization approach using deep neural networks, brighter AI is already on the way to solve the ‘data privacy vs. analytics dilemma’ for more than 1.1bn public cameras.

eCAPITAL is a venture capital firm that provides early to growth stage funding to technology companies in the fields of software & information technology, cybersecurity, industry 4.0, new materials and cleantech. Founded in 1999, eCAPITAL has a history of supporting entrepreneurs determined to build companies with lasting significance. Partnering with eCAPITAL means joining an international network of business leaders, entrepreneurs, technologists and potential partners. eCAPITAL is located in Germany, currently manages funds with over EUR 220 million under management and was lead investor of some very successful German exits like sonnen or Novaled.

Additional information and contact details
You can find a case study with additional information here.

Brighter AI Technologies GmbH
Marian Gläser, Co-Founder & CEO
Kopenhagener Straße 76
10437 Berlin, Germany
eMail: press@brighter.ai

Sylvia Richter
Hafenweg 24
48155 Münster, Germany
Tel.: +49 251 70 37 67 0
eMail: info@ecapital.vc