Certainly one of the most interesting aspects of the smartphone revolution in recent years has been the rise of personal assistants.  What began with Apple’s Siri has quickly spread, with Cortana (Microsoft) and Now (Google) all aiming to assist us in our daily lives.

That these technologies are relatively reliable is an achievement in human language processing in itself, but as with most things in the mobile world, the IP is fiercely protected for each service.

Step forward Sirius, a new mobile PA service developed by researchers at the University of Michigan that is built on open source principals.

“Now the core technology is out of the bag, and we all have access to it,” the researchers say. “Instead of making an app to run on the Apple Watch, for example, maybe I could make my own watch. We’re very excited to see what the world comes together to build and learn with Sirius as a starting point.”

The researchers launched Sirius at a recent technology conference in Istanbul, with the publication of a paper documenting their work, and the release of source code for the project.

The application utilizes the same speech recognition, image matching, natural language processing and Q&A systems as the existing PA services.  The initial launch will allow users to use voice and images to interact with the device.  The use of images is novel, as users can, for instance, upload a photo of a restaurant and ask what time it shuts.

“What we’ve done with Sirius is pushed the limits of the traditional intelligent personal assistant,” the researchers say. “Not only can you interact with your voice but you can also ask questions about what you’re seeing, which is a new way to interact with this type of device.”

Whilst the initial service utilizes a Wikipedia style of general knowledge, the researchers believe that this could easily be swapped for a more specific knowledge base for a particular niche.

For instance, a PA for medicine could be created or one for cooking.

The hope is that Sirius will be for PAs what Linux has been for operating systems.  Interestingly, the very construction of Sirius relied on open source approaches, with four established open source projects used as part of the project.

Speech recognition, for instance, was delivered via the Sphinx project frm Carnegie Mellon, whilst OpenEphyra helped to power the Q&A facility.

In addition to providing the next generation of PA systems, the researchers hope that the diversity of input into Sirius will greatly assist the research into other areas, such as the interaction between data warehousing and voice controls.