CUSTOMER SUCCESS STORY / HEILBRONN CITY ARCHIVE
City archive uses AI to digitise its processes

INDUSTRY
City Archive
PRODUCT
Fast and comprehensive recognition of historical materials
DeepVA helps the Heilbronn City Archive to automatically index and digitize a large amount of images.
Over 1,500 Heilbronn personalities and 200 buildings were trained and recognized with an astonishing accuracy.
Since 1371, the City Archive of Heilbronn has made a significant and relevant contribution to the remembrance of the region. It takes care of the city’s history research and serves as an exhibition venue and documentation center. It stores photos and newspapers, documents, files, maps and plans as well as literature on historical and local topics. All citizens interested in history can use the archive and its facilities, as well as all material from any accessible collection. Thus, the City Archive of Heilbronn promotes historical awareness and makes the historical treasure of the city accessible and tangible.
Challenge: Too many images and too little time to archive them properly
The archived and collected material of various types and origins is reviewed, evaluated and consolidated with the already stored material by the staff. Among them are about one million photos, the oldest from 1842, and every year this stock continues to grow in the four-digit range. So far, however, only about 50,000 photos have been labeled and included in the database. The amount of data to be digitized is large and time resources are scarce. In addition, the specialized knowledge needed to keyword these photos is closely linked to the archivists, and training new employees is complex and time-consuming.
The use of artificial intelligence in the city archive has met our expectations and provides real added value – not only for archival tasks, but also for interested citizens.
Miriam Eberlein
Project manager
Solution: automating the indexing of visual assets using the latest AI technology
In collaboration with the Freiburg-based startup The Chainless and its AI platform DeepVA, the Heilbronn City Archive was able to automate the keywording of photographic material and enrich large volumes of data with meta-information. The goal was to index more photos with the same amount of staff, make them searchable and implement them in the already existing IT infrastructure of the Heilbronn City Archive. To this end, more than 1,500 personalities and 200 buildings were trained. This goal was therefore more than fulfilled. A first quality check of the automatic keywording has shown that 92.5% of the persons are correctly recognized, as well as 75% of the buildings. The city archive is able to further improve the recognition accuracy independently through further AI training. DeepVA’s user-friendliness made it easy for the city archive to successfully adopt AI. In this way, city-specific knowledge can be preserved in the long term, independently of employees.
Result: The Heilbronn City Archive is the leader of municipal innovation
The cooperation between the Heilbronn City Archive and DeepVA received an award in the 2019 state competition “Future Communities” and was funded with 66,375 euros from the state of Baden-Württemberg. At the e-Government Competition 2020, the successful implementation was awarded second prize in the category “Best Project for the Use of Innovative Technologies”as one of the most trend-setting partnerships for more agility and efficiency in public administration. The city archive can now access significantly more photographic material with the same amount of staff and immensely improve the quality of archiving and research. Heilbronn is taking on a pioneering role in the digitization of the public sector and thus inspires other municipalities to introduce forward-looking technologies.
ACCURACY
NEED OF AI EXPERTS
FASTER LABELLING
Take a look at our other success stories

TVN
AI-powered subtitling for the German Chancellor ensures accurate, barrier-free broadcasts with real-time transcription, human editorial oversight, and GDPR compliance.

Arvato Systems
With the help of DeepVA, the user can intuitively and easily create and manage training data directly in the media asset management system (MAM).

Bayerischer Rundfunk
Creating and managing training data for face recognition using AI requires a lot of time and resources.

rbb
rbb’s collaboration with DeepVA transformed regional broadcasting and archiving through efficient AI utilization.