Can Robots Play a Role in Improving Lives of Autistic Individuals?

Lately I am noticing a lot of interest from the robotics community in developing robots to help autistic individuals.  Some of these efforts are based on technology push, i.e., people have developed a cool new robot and they would like to see if autistic individuals can benefit from using it.  Some efforts are genuinely targeted at understanding the needs of the autistic individuals and developing solutions to help them.

This post shares my thoughts on this topic based on our family’s experiences in raising an autistic daughter. Let me begin by setting the context.  Autism is a neurodevelopment disorder that affects the brain development.  Representative symptoms associated with autism include difficulty with social interaction, limited verbal and non-verbal communication, and repetitive behaviors. Individuals with autism face many challenges in their daily lives.

The intensity of symptoms associated with autism can vary from mild to very severe and there is a considerable variation in symptoms.  Experts use the term autism spectrum disorder (ASD) to refer to autism and related disorders.  It is often said that no two autistic individuals are alike. According to the recent statistics, one out of every eighty eight children born in the U.S. is diagnosed with ASD.  Unfortunately, there is no known medical cure for autism, making it a pressing social problem.

Individuals with ASD struggle every day to live in the world designed for neuro-typical individuals.  Most autistic individuals are hyper sensitive and often experience sensory overloads. Sounds, smells, and sights that might appear normal to most people often can overpower the senses of autistic individuals. They use stimulatory repetitive behaviors to compensate for the sensory overload.  Many autistic individuals struggle with language. They have a basic understanding of the vocabulary and grammar, but advanced language concepts are often foreign to them. Many autistic individuals are good in picking up body language cues from their peers and can sense the disapproval and rejection of their behaviors by their neuro-typical peers. However, most autistic individuals are helpless in controlling their behaviors; their own bodies and brains betray them every day.

Here is how the world might appear to an autistic teen as he/she goes through the daily life. Imagine that you are in the 10th grade science class.  The heating system in the class is making a really loud annoying thumping noise. This is crippling your ability to think. You try to cover your ears and start humming to drown that excruciating sound.  Your science teacher is delivering the science lecture in a “foreign language”. You understand the basics, but you are unable to follow the advanced vocabulary being used in the class. You are extremely frustrated and the stress is making it impossible for you to sit in your seat, so you are constantly fidgeting. You are noticing disapproving looks from your peers who find you weird and annoying. You are feeling humiliated and unwelcome in the class. You would like to fit in, but you are unable to control movements of your own body. The teacher has just announced that the next class will have a quiz. Quizzes make you really anxious and now you can feel a knot forming in your stomach.  Nausea has kicked in and the simple task of walking from Science classroom to English classroom appears to be a Herculean task.    

Unfortunately,  parents are often helpless and unable to eliminate the pain and suffering of their ASD children.  Providing care for autistic individuals can be emotionally and physically exhausting.   Most parents try very hard to improve lives of their children. Unfortunately, they also worry non-stop as to what will happen to their ASD sons and daughters as they grow old and unable to care for them.  Unfortunately, there is no good answer.  This can be a tiring, frustrating, and heart-breaking experience.  But this unfortunate adversity in life also showcases the resiliency of the human spirit. You meet so many individuals who do not give up and continue to fight incredibility hard to put one more smile on the faces of their loved ones and make the world a fair place by demanding universal accessibility.
  
Given this background, the question is - can robots play a role in improving the lives of autistic individuals?  We will have to approach this question very carefully as learning to interact with humans is a key to the survival of autistic individuals in the neuro-typical world. Robots should not try to reduce the human involvement in the lives of autistic individuals. However, robots can be useful in one of the following situations:

1. Increasing the human interaction will be detrimental to the intended outcome.
    
2. The use of robots can significantly improve the quality of life for autistic individuals.
    
3. Humans with the right expertise are not available to meet the needs of autistic individuals.

Here are my preliminary thoughts on potential applications of robots based on the above described situations.  

• Overcoming Positive Interaction Deficit: The human brain is wired to seek positive social interaction. Many autistic individuals also crave positive social interaction. However, it is very hard for them to interact with neuro-typical individuals and this can be quite frustrating for them. The lack of adequate amount of positive social interaction can lead to severe depression. In my opinion, there is no good way for us to overcome positive interaction deficit faced by autistic individuals by increasing the human interaction.  Human interaction is extremely important, but simply increasing the amount of human interaction does not mean that autistic individuals perceive this increase in a positive light. In fact, many autistic individuals prefer to interact with animals instead of humans because animals are non-judgmental and reciprocate affection unconditionally.  However, many autistic individuals are unable to take care of pets. I believe that robots can be designed to entertain, stimulate positive interaction, and uplift the moods of autistic individuals. Such robots must be carefully designed to ensure that they fulfill the positive interaction deficit and not try to replace the need for interacting with humans.       
   
• Improving Safety and Independence:  Many autistic individuals lack the basic notions of safety. This significantly worries caregivers and interferes with the freedom and personal space of autistic individuals. I believe that robotics-based technologies can be developed and adopted to enhance safety and independence of autistic individuals. These technologies can be used for safety monitoring (e.g., kitchen stove is switched off after use, medicine was taken on time), assist with household chores (e.g., cleaning), and navigation in complex surroundings (e.g., finding a store in a mall). There are many interesting technologies being developed for assisted living facilities that might find use in homes of autistic individuals. These technologies are not likely to look like a typical robot, but we should not care about the form.        
   
• Improving Training and Education:  We have to find a way to create meaningful employment opportunities for autistic individuals. Not doing this will create a significant financial strain on the rest of the society. Many autistic individuals have natural talents such as computers, music, and mathematics. These talents should be nurtured and harnessed. Learning to function well in the society will require developing appropriate social interaction skills such as making eye contact, reacting appropriately to facial expressions and body language, and making small talk. Currently autistic individuals get very limited opportunities to practice and hone these skills. Robots can be designed to enable autistic individuals to practice these skills for extended periods of time. It is very difficult to put effective special education teachers in all the classrooms with autistic children. Telepresence robots might be able to expand that geographic reach of superstar special education teachers and contribute to the training of autistic individuals.                   

I believe robots can play a useful role in improving the lives of autistic individuals, but we should take extreme care to ensure that robots do not displace humans from the lives of ASD individuals. Ultimately, human contact and interaction will be vital for ASD individuals to function well in the society.          

When can we buy robots to help us with household chores?

Recently Google has been in the news for its buying spree of robotics companies. Many people are excited about this and believe that this will greatly accelerate robotics technology development and hopefully make robots ubiquitous in our lives.

This post is mainly focused on robots for homes. In a simplistic sense, home robot market can be divided into the following three categories: (1) robots helping with dull and tedious household chores, (2) robots taking on new roles in homes (e.g., education, entertainment, companionship), and (3) robots in assisted living communities. Each of these markets has different underlying economics. In this post, I will focus on the first category.  Specifically, I am interested in the following question:  When can we buy robots to help us with household chores?

Before answering this question, let me quickly summarize the societal implications of home robots that can help with household chores:

• Home robots might save many marriages by reducing fights over household chores. I am sure that divorce lawyers will hate home robots!
• Teens will have love-hate relationship with robots. Parents won’t need to nag teens for doing chores. But teens won’t be able to make money by doing chores. Cash-deprived teens will need to cut down on money they spend on music and movies. This might be bad for certain pop stars. So watch out Justin Bieber, home robots won’t be good for your album sales.
• Home robots will dethrone pet animals from being the stars of viral YouTube videos. They might even spawn new reality shows on TV to depict new relationship dynamics at homes as new home robots join the family.  
• Call centers with human workers will be needed to bail out robots in distress. Hopefully, this will create new jobs for humans. Perhaps auto clubs like AAA can start new robot clubs to assist robot owners.
• People won’t have any sense of privacy inside their homes. Robots will be able to monitor your every move, so be careful of what you do at home. I am sure NSA folks will be very happy with the advent of home robots. They will finally have the ability to know what time you take showers. They already know everything else!   
• Occasionally, home robots will be involved in accidents. I can already see insurance companies salivating at this new opportunity and design new products. Hopefully, your robot can answer the phone when telemarketers from insurance company call you to sell new robot insurance policies.    

Here is a partial list of chores performed inside homes that can benefit from robot assistants:

• Help with laundry 
• Load and unload dishwashers 
• Do preparatory work for cooking meals (e.g., chopping vegetables)
• Clean kitchen
• Clean toilets and bathtubs
• Pick up objects (e.g., toys, newspaper, clothes, sneakers)  from the floor and move them to the right location 
• Unload groceries from the car parked in garage
• Assemble furniture
• Help with moving heavy objects
• Answer phone when telemarketers call at the dinner time

The above list does not include tasks for which the robot will have to venture outside the home. For example, I did not include lawn maintenance and snow cleaning tasks. A robot working outside the home will need to be able to deal with a wide variety of weather conditions and safety issues. This is a lot harder problem to solve.

I also on-purpose did not include pet sitting and baby sitting in the above list. Some people believe that robots should be able to do these. I think that this will be a good idea for TV shows, but a terrible idea in real-life.      

A home robot will require significant in-home installation, regular monitoring, and servicing to keep it operational. You may need to robot-proof your home to make sure that the robot does not damage your home and your home does not damage your robot.

I think we should look at a leasing or a renting model instead of a buying model. In this model, people will rent the robot from a company for a monthly fee.  The company will take care of the installation, monitoring, and service. If the robot is stuck, it should be able to contact a call center. Hopefully, someone in the call center should be able to teleoperate it to get out of the jam.

So how much people are willing to pay in monthly fees for robots at home? Based on my preliminary estimate, people will be willing to pay $200 to $500 per month for renting a robotic assistant for home.  I believe that people will be willing to pay $5 to $10 per hour of labor saved.  So if a robot can save 40 hours of tedious chores per month, then people will be willing to pay $200 to $400 per month for the robot.

Let us assume that a well-designed robot will have a service life of five years.  So the home robot market looks a lot like automobile market from pricing and service life points of view. 

In order for robots to be able to do forty hours’ worth of useful chores in homes, a lot of new technology will need to be developed. I believe that this technology can be sold at $200 per month if there were few million customers.  So here is the catch: unless there is a large market, the desired robots cannot be offered at the right price. However, unless the useful technology is available at the right price, the large market simply won’t exist.

Unfortunately, incremental development of home robot technology and its introduction in markets will be extremely slow. We will need to meet or exceed forty hours per month of useful robotic chores at home to create a significant home market and associated infrastructure. In my opinion, realizing a home robot is technologically feasible, but it will require billions of dollars of investment in technology development to ensure the high level of reliability and safety for home use. Unfortunately, venture capitalists don’t like these kinds of markets. My hope is that a cash-rich company such as Google, Microsoft, or Apple will go after developing this technology and create a new industry.  

Cell phones were invented for people to talk, but they have found new roles such as music players, web browsers, email clients, etc. The revenue growth in the cell phone market came because of the new roles played by cell phones.   I believe that the same thing is likely to happen for home robots as well. Initially people will be interested in getting robots at home to help with the household chores, but soon they will find new uses for these robots. I believe that robots might find an easier path to become fixtures in people’s home by adopting new roles such as tutors (e.g., music instructor, golf instructor), personal trainers, and entertainers.  

So, when can we buy robots to help us with household chores? If a cash-rich company like Google, Microsoft, or Apple goes after this technology, then we might have these robots before the end of this decade. Otherwise, we may need to wait for a while.

International Robot Exhibition 2013: Interesting Trends and their Implications

I attended the International Robot Exhibition (iREX) in Tokyo in November 2013.  It was a mesmerizing display of robots – a gigantic hall filled with thousands of robots.  Robotics companies bring their latest and greatest robots to this exhibition.  As you walked through the exhibition hall, you saw a wide variety of amazing advances in the field of robotics.

I noticed several common trends in new product offerings from many different companies. The underlying technologies behind these products were proposed many years ago, but for a while these were serving niche markets. However, it appears that suddenly these technologies have become mainstream, and several different large established companies are featuring new products based on these ideas. So finally after many years of wait, these ideas have moved from labs to the mainstream robotics industry. 

Here is my pick of four noteworthy trends based on products offered by established companies in robotics space: 
 
Dual Arm Robots: Humans (and many other primates) have two arms, but industrial robots for the longest time have featured only single arms. The argument was that if a task needed two arms, you can buy two arms and mount them next to each other.  The mainstream robotics companies resisted the idea of connecting two arms to a body and selling it as an integrated package. However, it appears that thinking in the industrial robots community has changed over the last couple of years. Many companies at iREX were displaying new robots with two arms. In my opinion, the dual arm robot configuration will provide new advances in the dexterous manipulation area where two arms can be moved in a coordinated way to work with complex tools. Humans have a naturally tendency to utilize both of their hands when doing a task. Imagine cooking dinner with one hand tied behind your back! So dual arm configuration should it make it much easier for humans and robots to collaborate on complex tasks.    

ABB Dual Arm Robot

Nachi Dual Arm Robot

Eyes on the Hand: I saw several robots with cameras mounted very close to the hand. This configuration gives robots unobstructed close-up view of the parts being manipulated. This idea was proposed more than twenty years ago, but there were reservations in implementing it on the shop floor due to concerns about acquiring quality images and registering the images with a fast moving camera. I am happy to see that these challenges have been overcome and this configuration is featured on many robots. This configuration will enable new advances in visual servoing and enhance the accuracy of the fine manipulation of objects previously unseen by the robot. It is interesting to note that in the first trend reported above, companies created robots that embraced the anthropomorphic configuration. However, this trend moved robots away from anthropomorphic configuration by placing eyes on the hand. Cameras are inexpensive, so robots can afford to have eyes on the limbs. I am sure that many humans have wished that they had a pair of extra eyes.

Motoman Robot with Camera on Hand

Wearable Robots: There were many different kinds of robots on display that people can wear to enhance their capabilities, ranging from walking assist devices to exoskeletons. Some of these robots are targeting the physical therapy and rehabilitation market to help people recover from injuries or loss of motor functions due to medical complications (e.g., stroke). Some robots are targeting the assistive technology market to help people cope with diminished abilities due to aging or other medical conditions. It appears that the robotics industry has combined high efficiency actuators, lightweight structural materials, and new battery technologies to finally create useful products. Wearable robots are expected to positively impact the quality of life as the average human lifespan continues to increase due to the advances in medicine. They also provide new ways to carry out physical therapy and rehabilitation. I believe that they will eventually enter the sports market to help with athlete training. There is plenty of room in the amateur market too. It will be great to have a wearable robot that can teach you how to swing your golf club.

Honda Walking Assist Device

High Speed Pick and Place Robots Based on Parallel Kinematics:  Parallel kinematics based robots hold significant promise because the actuators can be placed near the base of the robot, significantly reducing the inertia of the moving links and enabling high speed operation. I was happy to see that every major company was featuring high speed pick and place robots based on parallel kinematics. Companies were reporting impressive workspace sizes, high repeatability, and large payload capacity in robots based on parallel kinematics. These robots are bringing speeds comparable to the hardware based fixed automation to programmable automation.               

ABB Flex Picker

Kawasaki Delta Robot

The Role of Robots in Engineering Education

Currently, the robotics industry does not have enough jobs to employ all the engineering graduates who would like to pursue a career in robotics. Sometimes this leads to disappointment among engineering graduates who have passionately pursued robotics in school and find it frustrating that they cannot find a job in the robotics industry.  Hopefully, the robotics industry will continue to grow and the job situation will be significantly better a few years down the road. But what should graduating students interested in robotics do in the meantime?

This blog post attempts to argue that students should use robots as learning tools to acquire a much broader engineering knowledge base to impress employers in a wide variety of engineering industries. 

Human beings are fascinated by robots. I don’t fully understand the reason for this fascination, but robots seem to get people excited and enthralled.  In addition to being popular inhabitants of our factories, hospitals, and farms, robots are also emerging as popular cultural icons with an ever-increasing presence in music, movies, and books. Clearly the idea of humans ultimately creating superhuman robots is quite thrilling and intriguing. Let us also keep in mind that robots can be quite entertaining. 

Robots have become ambassadors of STEM education in the US. Many K-12 students get their first glimpses of the engineering world by participating in FIRST robotics.  Children as young as five years old are getting started in Junior FIRST LEGO Leagues, and they can continue advancing through various FIRST programs as they gain experience and grow older. The FIRST robotics experience finally culminates into the FIRST Robotics Competition.  In this program, teams of high school students compete with each other in a high profile national competition by building incredibly impressive robots. More than 350,000 students are expected to participate in the FIRST programs in 2013 and 2014 season.

Many undergraduate students who enter college with FIRST experience in high school take robotics courses in the college, continue their participation in challenging robotics competitions, and go on to build even more impressive robots. Some of the students interested in robotics go on to graduate school and continue their journey in the field of robotics. National Robotics Initiative and DARPA Robotics Challenge are providing many new exciting opportunities for students in the US.  

Along the way some students start to focus too much on “learning robotics.” It will be better from the employment perspective, if students viewed robots as a vehicle to learn engineering. The FIRST programs do a great job of emphasizing this point, but somehow along the way this message gets garbled for at least some students and the focus shifts to learning specialized tools for building robots.

Robots are very good examples of modern cyber physical systems. They contain mechanical, electrical, electronic, and software components that interact in complex ways to produce the desired behavior and performance. In terms of building blocks, they are no different than modern dish washers, automobiles, magnetic resonance imaging machines, and cranes. However, robots are a lot more fun to create than a dish washer and hence much more effective for teaching engineering principles.  

We should view robots as tools for learning modern engineering principles and have fun while doing it! Once you have learned how to design, build, and program a robot, you can create lots of other engineered artifacts that people are willing to buy today. Your experience with robots makes you an excellent catch in the job market. You should just learn to interpret the word robotics in the broadest possible sense.

Can crowdsourcing be exploited to reduce the cost of autonomous robots?

Several recently developed prototypes show that it is possible to develop autonomous robots with remarkable capabilities. However, currently developing autonomous robots costs a fortune and takes forever!  In this post, I treat autonomous cars and unmanned air vehicles as robots.

Sophisticated autonomous robots require hundreds of thousands of lines of code. Manually writing this code for a new robot is very expensive and time consuming. Moreover, as the hardware changes, this code also requires significant upgrades. Often by the time code is written and debugged, the hardware is already obsolete. Therefore, developing autonomous robots is currently technically feasible but not affordable in many applications.

Human operators are very good in teleoperating robots in cluttered, unstructured, and dynamic environments with limited sensor data. Forget about the expert operators teleoperating unmanned vehicles! Even five year olds can learn to teleoperate their first remote control cars within couple of hours and are able to successfully annoy their parents, siblings, and dogs by zipping around tiny cars inside their homes. I have also seen teenagers performing amazing feats with their remotely controlled helicopters. So obviously, we should be interested in characterizing and understanding the strategies employed by human operators during these operations and automatically extracting building blocks of the autonomy code based on this understanding. Many robotics researchers are pursuing this path and this area of robotics is called learning from demonstrations.

Many impressive results ranging from training of surgical robots to teaching collision avoidance to unmanned vehicles have been reported by the learning from demonstrations community. Most such case studies have utilized small number of humans to perform these demonstrations. Because of the limited number of demonstrations, people often wonder how well the learned components of autonomy will perform in situations not encountered during demonstrations. Unfortunately, conducting extensive experiments in physical worlds is highly time consuming and expensive. It also limits the kind of scenarios that can be considered during demonstrations. Clearly a demonstration that might pose threat to the human or the robot has to be avoided. Conducting demonstrations in the virtual world is emerging as an attractive alternative.

Over the last few years, tremendous progress has been made in the area of physics-based robot simulators. For example, the on-going DARPA Robotics Challenge is making an extensive use of simulation technology to test autonomy components. Simulations are being routinely used to teach humans cognitive as well as motor skills. For example, flight simulators are routinely used for pilot training.

By combining advances in multi-player games that can be played over the network and accurate robot simulations, new games can be developed in which humans can compete and collaborate with each other by teleoperating virtual robots. This advancement means that demonstrations need not be confined to few experts. Instead, anyone with an Internet connection can participate in the training of a new robot. For example, DARPA used publicly distributed Anti-Submarine Warfare game to learn how to track quiet submarines. We are ready to leverage crowds to impart autonomy to robots. 

The use of crowd sourcing in robot training has many benefits. It provides a rich diversity in demonstrations and hence enhances the probability of generalization. Some of the participants are likely to exhibit out-of-the-box thinking and demonstrate a highly creative or innovative way of doing a task. This is great news for optimizing robot performance. For some people, this way of training robots might serve as a means to earn money by performing demonstrations (basically acting as robot tutors). Playing games that involve robots is likely to be entertaining for at least a segment of the population. This paradigm can also be used in situations where a robot is stuck during a difficult task and needs a creative solution to get out of the bind.               

Automatically learning autonomy components such as reasoning rules, controllers, planners etc. from the vast amount of demonstration data is an interesting challenge and will keep the research community busy for many years to come. But this seems to be the much needed crucial advancement to reduce the cost of autonomous robots.

Don’t worry robots! The crowd will rescue you from the dungeons of high cost and long development times!