Verity is the global leader in autonomous indoor drone systems. Our integrated inventory tracking system enables zero-error warehousing for retailers, third-party logistics providers, and manufacturers.

People generally use addresses to communicate points of interest, such as streets and numbers for buildings, aisle numbers in supermarkets, or room numberings inside buildings. People routinely use these addressing schemes to find places previously unknown to them, but machines struggle to do so without a complete address book.

The goal of this master thesis is to investigate estimation methods that infer an address book based on imperfect sparse observations of address-location pairs and logical addressing rules. As an extension, optimal observation sampling methods can be investigated. Relevant topics include:

• Machine learning

• Bayesian inference

• Factor graphs

This thesis will be carried out at Verity AG, where fleets of autonomous drones will use the inferred address book to find their way through some of the largest warehouses in the world. Verity's system is used by industry leaders around the world, including DSV, Maersk, and Kehe. Verity's live events solutions help create stunning effects for top names in the entertainment industry like Drake, Céline Dion, Justin Bieber, Metallica, and Cirque du Soleil. We continuously reach beyond existing limitations to re-think, reimagine, and rework the role of robotics in the world today.