San Diego is a joint venture between University of California-San Diego and a Japanese robot manufacturer, Kokoro. It was created to study how human infants learn, grow and interact with the world. (In other words, they wanted to study toddlers without actually having to deal with toddlers.)
San Diego cannot speak, but it can gesture and offer facial expressions. The toddler-bot has about 60 moving parts, about 20 of which are in its face, allowing for a wide spectrum of expression. It can also can get up from a chair and grasp objects such as plastic bottles. Baby Diego also sports high-resolution eye cameras, ear sensors that can detect movement, and a speaker in its mouth. Pressure sensors can detect the load on its joints. Apparently, the robot's gigantic head is necessary to accommodate all his electronic parts and computer chips. It is just over 4 feet tall and weighs 66 pounds. It's about 1.5 times the size of a human baby.
So, what am I missing here? Can someone explain to me, why would you go through all of the trouble and expense of building this odd-looking robot to study how human babies "learn, grow and interact with the world," when this cyber-tot's abilities to do just those things are at the mercy of its programming? If you want this little fella to put his toys away, you just program him to put his toys away. Want him to read? Just write the program. Hmm, on second thought that does sound appealing. Nah ... while this pre-programmed robo-toddler may sound ideal to some, I'll take a regular old skin-covered, diaper-bottomed, tantrum-throwing toddler over this robot creation any day.