Microsoft's Giving Us A Look At The Future Of Robots In The Office
Coffee in hand, you stagger into your workplace, making for the elevator. Detecting your presence, it automatically opens its doors as you approach, inviting you in. As you step through, a voice echoes through an embedded speaker system:
"Same floor as always? "
You offer a bleary nod, and the doors close as the elevator carries you to your destination. After arriving at your floor, you walk to your desk, where an assistant approaches you with your day already planned out. You're informed of a pending lunch meeting, and of several appointments that need to be rescheduled. You voice your approval and leave the details up to your secretary. He knows you better than you know yourself, sometimes.
Have I mentioned that all of this happens without ever having to speak to another human being?
These two features are just a few components of Microsoft's Situated Interaction Project, an immersive experiment designed by scientists Eric Horvitz and Dan Bohus. Designed as much to showcase the future of human-computer interaction as it is to test and develop new technologies, Situated Interaction takes the form of a fully-automated office in which robots and humans work in tandem. With the project, Microsoft also hopes to demonstrate that humans and robots need not take an adversarial relationship in the workplace; that the two working together can accomplish far more than they could on their own.
The fully-interactive experience brings together an intensive array of computing competencies, methods, and technologies; including machine vision, natural language processing, automated planning, speech recognition, acoustical analysis, machine learning, and sociolinguistics. More interestingly, it represents Microsoft's push into a new area of research, as well: the automation of processes and systems designed for multiparty interaction, which has always been a core challenge of artificial intelligence.
"The goal," explains Horvitz, "is to build systems that can coordinate and collaborate with people in a fluid, natural manner. Intelligent, supportive assistants that assist and complement people are a key aspiration in computer science, which is why we expect fierce competition from similar companies in the same space."
For example, Horvitz's digital assistant has full access to his online calendar, and can detect if he's in the office. It's capable of figuring out how busy he is by studying him, and can even predict when he'll finish a particular task or conversation based on past habits it's observed. In other words, it's capable of connecting with him in much the same way as a real assistant would.
According to Microsoft, the applications for the technology here are nearly endless, and could extend far beyond the office world. Assistive robotics could end up serving purposes as wide-ranging as medicine, aerospace, and disaster relief. Although the focus at the moment is on making mobile devices more intuitive, it's pretty clear what the future holds.
Far from replacing people with robots - as some seem to suggest Situated Interaction will ultimately do - my perspective is that it's designed to show why we shouldn't do just that. Rather than designing fully-mechanical workplaces in which a human presence is wholly unnecessary, there's far more to gain by creating spaces in which robots and humans can function as two parts of the same whole. That, I think, is the purpose of Situated Interaction, and the future of robots and artificial intelligence in the business world.