Founded in 2019, ETH spin-off, Animatico, has developed software for interactive, talking avatars. The animated characters can provide support in various areas of everyday life, such as buying tickets, checking into hotels, or advising on products. The systems work without any contact – certainly needed in Covid-19 times.
A conversation with Animatico co-founder, Christian Schüller.
Christian, how did you get the idea for launching a software start-up with talking avatars?
During my studies in Computer Sciences and subsequent doctorate at ETH Zurich, Pascal Bérard, Riccardo Roveri, and I worked intensively on technologies in the field of computer graphics and computer-aided learning. Initially, technical motivations were the primary focus – bringing digital beings to life and making them tangible for everyone. Thanks to our previous work at ETH and at Disney Research, we knew the potential these beings had in animated films. In everyday life, we increasingly find ourselves in situations in which we see the customer experience – on, for example, a complicated ticket machine – as frustrating. So we thought, “That can be done better”. This gave rise to the idea of avatars as user interfaces for digital devices that people can also speak to and interact with on an emotional level.
What are the advantages of avatars? In which areas can the software be used?
Interacting with the avatar is very easy; it feels natural and trusting. There’s a lot of development behind this. Through voice, gestures, emotions, and visual representations, the avatars support and guide users step by step through a specific concern. In this way, avatars enable simplified and completely intuitive device operation in the area of customer service, such as ticket-buying, hotel check-in, or advising on products. For example, if a person is kindly asked by a concierge avatar in a hotel uniform if they need a taxi or need help booking a restaurant, people can use language to find what they’re looking for faster than with buttons or a virtual keyboard. Basically, the system can either operate completely without having to touch a screen, but also offers the classic touch operation for support.
Are there comparable systems?
Our key technologies – with the help of a small camera – are the dynamic visualisation of the avatars and the recognition of user interactions. This makes us completely different from language-based assistants that need a “wake word” (predefined word to activate an interactive application) to activate, and often have to cover a wide range of possible questions, which – among other things – leads to longer response times. Our avatars can be seen as a software variant of “social robots” (physical robots that interact with people). Physical robots are often very expensive to manufacture, have a predefined appearance, require a space in which they can move and therefore require regular maintenance. Animatico can visually adapt its avatars quickly for every conceivable application. The software works on every common screen and tablet. The Animatico solution is very easy to scale and is particularly suitable for use-cases with a large number of interaction points.
Who are your customers, and do you already have collaborations? Can one “experience” Animatico anywhere?
While we are working with several companies on specific pilot projects, these aren’t yet public. Our customers are companies from all sectors including healthcare, hotels, retirement homes, finance and insurance, mobility, entertainment such as casinos, and many more. Thanks to the software, companies can improve existing kiosk solutions, and break new ground in customer service by optimising and making it more efficient.
What are your plans for 2020/2021?
We’re right in the middle of the “seed” financing round that we want to use towards bringing our current avatar software to market maturity and to implementing new customer projects. Since we want to expand our current three-person team soon, we’re still looking for support in software development and on the business side.