background image

These aren’t the droids you’re looking for

I recently spoke at InnoRobo/RoboLift, a European robotics conference, held in the beautiful city of Lyon, France. Since my last trip to a European robotics conference was about 3 years ago, I was looking forward to catching up on how robotics was developing in this part of the world. One of the first things I noted was that the conference organizers had invited speakers from the design and user research community including Cynthia Breazeal, professor from MIT’s Media Lab. I particularly like to hear about Cynthia’s research because I share her philosophy that robot interaction needs to be designed as a social experience. Her work with Kismet, Leonardo, Nexi, and other robots really stand out as great examples of where the design of personal robots needs to go. This was refreshing, because many robotics conferences I have attended (except for the annual HRI conference) typically focus on optimistic market projections or demonstrations of the latest software algorithms or hardware. It is a good sign that the emerging industry is starting to recognize the importance of the role of design.

This conference still had its share of technology demonstrations. Aldebaran’s Nao robot seemed to attract the most attention from the press and attendees. This small walking humanoid robot, like a mini-Asimo, continues to fascinate the crowds, taking over the stage from the ill-fated Sony Aibo that used to be a star attraction. However, as impressive as the Nao is, it still is fairly expensive, roughly $16,000. And while it can be programmed to do a number of interesting tricks, like Aibo, is  primarily a research platform and like Aibo fails to deliver a strong enough consumer value proposition that  justifies its cost. Aldebaran has also promised a larger version, as part of Romeo, a 10M Euro, French research collaboration, with the goal of assisting elderly and disabled users. However, having an estimated initial price point of 250,000 Euros, this robot like the Honda Asimo, seems to be out of practical affordability for users anytime in the near future. And like Asimo, while the Aldebaran Romeo humanoid form implies a wide range of possible functions, it still leaves open many questions of what value it will be able to deliver that will justify its price. Robots have succeeded in the industrial space because they are programmed to do repetitive operations within a confined space. Trying to do this within a domestic environment and do common household tasks is many orders of magnitude harder.

Further, like so many robots of this type, Nao is one more example that implies that we will be able to just talk to our personal robots. I wish that could be true, but anyone familiar with speech technology knows that as a primary modality speech input fails. Even human speech recognition has only about 70% accuracy, which means that it fails about 30% of the time. If your keyboard or mouse failed this often it is unlikely you would use them. Speech works in human dialogues because we use other contextual clues, including things like our understanding of syntax and grammar as well as semantic associations. These have not been sufficiently replicated so far (see my previous blog entry Elementary, My Dear Mr. Watson for further comments on this.)

In addition to Nao, there were many other impressive robots demonstrated at this event in a variety of forms, from robot dogs and fish to robot toys. Overall, from a technological perspective, it was a very impressive and varied array of bots that reinforces that the robots and the components to create them are increasingly affordable and available in a variety of forms. Yet nothing here demonstrated that we are there yet. While this conference also included its share of speakers who enthusiastically presented optimistic market forecasts for robots, I saw little at the event that supported that we have progressed very far in having robots in our homes beyond the toys and gadgets of today. Colin Angle, CEO of iRobot, in a more balanced presentation, noted that if the industry’s rosy optimism is to be fulfilled, companies need to focus more on creating value to users. I agree and similarly addressed this in my own presentation, adding the need for more attention on the design of the user experience.

Speaking of  iRobot, I was impressed with their introduction of Ava at CES early this year and again more recently demoed at Google I/O. Ava incorporates many interesting features including an omni/holonomic drive that enables the robot to move in any direction, a torso that can adjust its height based on user touch-contact, and a mount for a touch-tablet (iPad or other). Using a laser sensor and a pair of depth cameras iRobot indicates that they plan to include built-in support for autonomous navigation, enabling developers focus on applications. However, as of this date, pricing and available are not yet known. And while it is a very nice platform it still does not put forward a specific design for human-robot interaction or yet what the robots actually will do that provides value to the user. Similarly, RoboDynamics, who has  worked on several remote presence robots, unveiled their new Luna robot, also sporting a single depth camera, but apparently not providing any navigation software. Their reported price of around $3,000 is attracting attention. It looks like depth cameras are increasingly popular sensors for robots as they are also featured in the new Willow Garage Turtle robot.

It is great to see all these platforms emerging, but I still believe that the key catalyst for requires more than offering an innovative platform, set of enabling technologies, or even an app store. Tablets existed before the iPad. Smartphones existed before the iPhone. What is still missing from all these robots is a core set of applications that provide clearly identifiable user value as well as a “magical” user experience. While Apple’s success greatly benefited from creating their App Store, the initial success of iPhone was based on its delivery of a useful suite of applications, so that the device delivered value at the right price on day one, wrapped in a user interface that people found more appealing that other phones. Clearly having a successful developer platform and a mechanism to distributing the creative add-ons helped, but this was something they were able to leverage by first building a compelling foundation.

So perhaps with these recent announcements we are getting closer, but we are still missing demonstration of key factors that will transform the industry from toys and gadgets to a new consumer category.  While innovation of hardware is great, delivery of user value  at an appropriate price and user-centered design even more needed. They are the prime directive at Hoaloha Robotics.