RoboBusiness Conference: Too Many Robotic Hammers Looking for a Nail?

Last week’s RoboBusiness conference in Boston was chicken soup for the geek’s soul. For the inner kid inside all of us, it showed how all that we have imagined over the past 50 years in books and movies can become a reality. Star Wars, Transformers, Terminator, it’s almost all possible now through advances in robotics technology. However, for the outer business person in us, the RoboBusiness show left you thinking: “What’s wrong? Why isn’t the Robotics industry ahead of where it is? Why don’t we have wide-scale adoption of consumer robotics? What’s slowing things down?”

While several sectors of the robotics industry have seen tremendous success, namely industrial and manufacturing robotics, consumer robotics is a hodge-podge of expensive toys, hobby-kits, and very sophisticated (and expensive) personal robots. Even bright shining stars like iRobot struggled during their early years, in iRobot’s case taking 12 years to achieve a consumer break-through with the launch of the Roomba while also tapping into the lucrative field of military robotics. While that story has a happy ending, for every iRobot there are ten other companies that burned through investors’ money and found their way to the scrap-heap of robotics history. Why is it that, where other technology industries typically take 2-5 years to commercialize a viable product (or fail trying), in the consumer robotics world it can take a decade-plus to build a winner?

I had a chance to attend several of the conference sessions and, during the Q&As and afterwards during the cocktail reception, had a chance to pose these same questions to some of the industry leaders. I received a mix of puzzled affirmation and some compelling theories. Below is a summary of some of these conversations. Trender Research is also working on a detailed “Robotics Thought Leaders Report” that will capture words of wisdom and best practices from the industry's top experts (please email Brian Mahony bmahony@trenderresearch.com to nominate a thought leader for this report or if you wish to purchase a completed copy).

My journey began with a talk about “Assistive Robots” by Holly Yanco, Associate Professor, Computer Science Department, UMASS Lowell. In her session, she gave a brief history of different assistive robots as well as her assessment as to why many of them so far have come up short. One of the stories that stuck with me was an attempt to build a robotics nursing assistant that could, among other things, guide patients to their appointments within the hospital. The product was a case study in technology created without really thinking about how the product would interact with the user. The robo-nurse would ask questions like: “Are you still here?” but it did not use speech recognition to hear the response from the human, instead requiring tactile input on a screen. Bizarrely, this screen was on the front of the robot and could not easily be accessed as the robot led the human throughout the hospital. It also had no detector as to whether the human was keeping up, so the default for the robot was a snail’s pace, leaving the human frustrated and feeling like the robot was blocking the way. This story leaves us with a potential answer: does the robotics industry need to do a better job with focus groups, pilots, and other forms of user research? Is lack of good primary research and product management the bane of the industry?

Moving on from there, I sat in on a presentation by Jim Wyatt, Director at Kablamm, the maker of the very versatile Mech RC robot, which impressed me with its versatility and programmability when I saw it in action at CES earlier this year. Jim seems like a very practical guy and made a point that, while we are still in the early stages of the robot industry, it is very important for each robot to do what it is intended to do. For example, the Mech RC fits nicely into the “edutainment market.” It is a favorite of high school and college robot clubs and often used in robot competitions and events. Since it is very programmable and flexible—it can be taught dance moves for example— it has “legs” beyond the capabilities you get when you first bring it home. This programmability is an important quality in robotics, since it gives the robot a dynamic of evolving utility and also introduces a level of social interaction with other robot enthusiasts. One of the big problems for most robots, especially those designed as toys or companions, is that users get bored with them once the impulsive purchase and first use period is over. A robot that is programmable, and allows one to share ideas and interact with other humans, ties into a deep-rooted need for humans to grow and learn and feel like they are part of a group. So is increased programmability and robotic intelligence the answer to the consumer robotics industry’s barrier to explosive growth?

My next stop was a session by Robotics Trends President Dan Kara. He had an excellent presentation entitled “Selection Criteria for Consumer Robotics Products.” He did a great job breaking down the various robotics industry sub-categories and needs. For consumer robots, he listed the following variables as the top selection criteria: 1) Works as advertised (doesn’t underperform expectations); 2) Robust/Durable (doesn’t break); 3) Practical/solves a real problem; 4) Good customer support; 5) Price. While the ordinal ranking of “price” varies by sub-category of robot (e.g., toys, hobby-kits, personal assistants, etc.), it is interesting to note how these first three variables work together to summarize a universal rule for robotics product development: “Robots must not create more work or problems than they solve.” Outside of simple toy robots, a robot must be worth the money in terms of time and hassle saved. An example of a robot that is good but could be a lot better is iRobot’s Roomba. As I have pointed out before, while I generally like the Roomba, the problem I have with it is its small cargo capacity for dirt. Unlike a larger vacuum which might need to be emptied every few weeks or so, the Roomba needs to be emptied every day or so, a messy job that also becomes a psychological block for continued usage. This challenge significantly detracts from the value of a “hands-free” device that can be scheduled to automatically clean and then return to its home base for recharging (I am still waiting for iRobot to respond to my idea of the “Burping Roomba” which could regurgitate its dirt payload into a newly designed home base).

While Dan’s research did a great job outlining what consumers are most looking for, it did not really answer the question of what the industry must do achieve greater success in launching various robots to the industry. One of the problems in the robotics industry is the repeated attempts by robotics manufacturers to create a new robot and then look for applications and industries who might be able to use it, the perennial “hammer looking for nail” problem. It raises the question whether the robotics industry as a whole would be better organized as one big R&D shop with expertise in all the major robot technologies. In this way, industries and companies that best know what problems they have could come to them to look for ways to automate and systematize these processes through robotics.

This is precisely the conversation I had with Richard Mahoney (no relation), Director of the Robotics Program with industry giant SRI International. This is exactly what SRI does, so I figured Richard would fully agree with my proposition. But while he obviously touted the capabilities of an R&D partner like SRI, he was more optimistic about the ability of individual robotics companies to innovate in various industries where robotics can help. He highlighted the manufacturing industry as one where robots have seen wide-scale acceptance and demonstrated value. The correct model, he pointed out, could be a partnership between individual robotics companies, industry players, and engineering labs like his own, each party aware of its capabilities and expertise.

Perhaps the most forthright of the robotics experts at RoboBusiness was Matt Fischer, CEO of KumoTek, and start-up developer of hobby and entertainment robots. He lamented that the history of robot development has been a “top down” approach instead of a true “bottoms up” assessment of consumer needs. He gave an example of a $1,400 robotic MP3 player (speakers on robotic wheels) as a bad idea that was probably built for the “coolness factor” alone. In Matt’s view, the industry needs to “cross the chasm” by taking a practical approach. According to him, unless a robot “saves lives, is government funded, or continuously makes/saves consumers money”, it must follow these five major rules: 1) It must cost less than $300; 2) It must be designed around software rather than hardware capabilities like movement (meaning it must take into consideration how it interacts with users); 3) It must be aesthetically appealing; 4) It must be culturally appealing; and 5) It must have upgradability to provide continuing utility. Matt points to Hanson Robotics “Zeno” humanoid robot as one that potentially follows these rules. The consumer version of Zeno will cost less than $300, it is jointly developed in the U.S. and Japan to ensure wide appeal, and it embraces “neural networks” as a way to build programmability, intelligence, and social networking into the robot. Matt’s rules suggest that the robotics industry may need more MBA types to help create and refine the business plans for new robotics products. It struck me that most of the leaders of the robotics companies I met at RoboBusiness came up from the engineering side of the house.

While there may be no secret formula for success in consumer robotics, the RoboBusiness conference demonstrated that the potential in this industry is vast. To further explore what may be holding the robotics industry back from wide-scale consumer adoption, Trender Research is authoring a “Robotics Thought Leaders Report” that will capture the wisdom and best practices from the industry’s top experts. To nominate a thought leader for this report or to purchase a completed copy, please email Brian Mahony bmahony@trenderresearch.com.