AI and robotics are becoming sophisticated enough that they can help solve world problems such as hunger, alternative energy, disease, and crime. But we need to be realistic about the work still needed to develop them.
Listen to this article
|
Robots, drones, and artificial intelligence are evolving so rapidly, it’s natural to expect them to be powerful tools in addressing world problems. Challenges for researchers, vendors, and end users of automation include unrealistic expectations, impatience with the pace of important research, and a proliferating number of applications.
Robotics Business Review has partnered with Abishur Prakash at Center for Innovating the Future to provide its readers with cutting-edge insights into recent developments in international robotics, AI, and unmanned systems. Are you ready to be updated?
Robots could help address food insecurity
Robotics development: Scientists at the Tomsk Polytechnic University (TPU), a university in Russia, are among the teams developing robotic bees, or “robobees,” that can pollinate plants as effectively as real bees.
TPU plans to produce 100 of these robobees at a cost of $1.4 million and to test them in 2019. For now, the robobees will only be used in enclosed spaces, such as for strawberry plants. The university said it expects such robots to be tireless.
Geopolitical significance: Food security has always been a complex geopolitical challenge for countries. Lack of food can lead to social unrest, economic recession and even war — just take a look at what’s happening in Venezuela.
To solve world problems such as “food insecurity,” all kinds of robots are emerging. The robobee from Russia isn’t the first of its kind. Last year, a group of researchers at Japan’s National Institute of Advanced Industrial Science and Technology unveiled drones that can pollinate plants.
At the time, the drones were remotely controlled by people but the researchers said they were working on an autonomous version.
Entire farms are also being automated through robots. In the Hands Free Hectare (HFHa) project, a joint effort between Harper Adams University and Precision Decisions, entire farms have been automated. So far, HFHa has deployed robots to two farms to grow winter wheat.
It isn’t just physical robots that are being developed to grow food, either. There’s also AI. For instance, MIT has created a “food computer” that uses AI to assess the conditions that crops grow in and adjust the conditions to maximize growth and yields.
In one experiment with basil, the food computer kept on learning from different batches. When the computer was tasked with increasing the flavor, the AI controlled the light settings to increase the production of one flavor molecule by 895%.
As governments look at new ways to protect themselves from world problems with the food supply related to climate change, higher crop prices, or growing populations, robots may become the most important farmers of tomorrow.
The big question, though, is whether robots can grow food fast enough to keep up with the rising demand. According to the United Nations, food production has to rise 70% to feed the 9.6 billion people who’ll be living around the world in 2050.
AI to tackle the biggest world problems
Robotics development: In a joint project between the U.S. Department of Energy’s Princeton Plasma Physics Laboratory (PPPL) and Princeton University, a research team is looking to a supercomputer to tell them to how burn plasma fusion systems without disruptions. The results could move practical fusion energy closer to reality.
The project will use the Aurora supercomputer, which will be set up in 2021 and will be the first exascale system in the US.
Geopolitical significance: The demand for and distribution of energy resources is one of the biggest world problems. But creating viable energy sources beyond oil, natural gas, and coal has been difficult. The idea of fusion energy has remained just that — an idea.
If robots can help turn this idea into reality, then it will be a huge milestone in solving this world problem. Consider that fusion energy doesn’t emit greenhouse gases, is completely clean, and is self-dependent (it keeps itself heated, and excess heat is turned into electricity).
If robots can solve fusion energy, what about other world problems? Take the spread of deadly diseases such as HIV/AIDS.
At the end of 2016, an estimated 36.7 million people were living with AIDS. One of the largest groups at risk is homeless people; 7% of all homeless youth in the U.S. may be HIV-positive.
To solve this, the USC Center for AI for Society (CAIS) developed a system to inform the public better about HIV/AIDS. The AI system was 150% more effective than humans.
Machine learning and deep learning can also help design brand-new drugs and design treatments for other health problems, such as cancer and dementia.
Another set of world problems where automation can help is with fresh water. While AI hasn’t yet reached a point where it can find new fresh water or new ways to create fresh water, it is able to reduce water waste and improve water plants.
For example, the University of Waterloo has developed an AI that can scan water and identify when there is blue-green algae, a type of bacteria that causes water treatment plants to shut down because of its ability to spread rapidly.
Currently, water treatment plants send small samples of water to labs to be manually tested. Solving world problems may sound like a grand task, but AI is slowly inching towards a place where it may be able to do just that. General artificial intelligence promises to be up to the task.
Robotics and AI companies that can address the world problems of our time, such as water shortage, food security or disease spread, could receive serious interest and investments from governments.
Facial recognition can catch criminals
Robotics development: A facial recognition system at the Washington Dulles International Airport has caught its first criminal: a man from Brazil using someone else’s passport.
When the 26-year old approached the border agent at the airport and presented the passport, the border agent ran the passport’s photo through the facial recognition system. The system ran the photo against the man’s face and said it did not match. When the man was searched, an ID card from the Republic of Congo was discovered.
Geopolitical significance: Domestic security is the main way for nations to address world problems around domestic security. And facial recognition is enabling this by catching people who pose a risk.
A similar arrest to the one at Washington’s Dulles International Airport took place in China. In April, at a concert in Nanchang, China, facial recognition systems found a wanted man and alerted police. The police then arrived at the concert and arrested him.
What makes this so important is this: The man was at a concert with 60,000 other people, and the facial recognition system was able to differentiate and identify him.
Last year, police in Wales used automatic facial recognition (AFR) software to identify and arrest a man. It was the first time facial recognition resulted in an arrest in the U.K.
This past July, during the Russia World Cup, police in Moscow tapped a facial recognition system to arrest a soccer fan. The system was connected to the 160,000 CCTV cameras that make up Moscow’s “Safe City” surveillance system. It flagged a person attending the Spain-Russia game.
However, facial recognition has its wrinkles. In July, Amazon’s facial recognition system incorrectly identified 28 members of Congress as criminals.
In the same month, the U.K.’s Scotland Yard reported that, during a facial recognition trial, not a single arrest was made. The system did not identify anyone as a criminal, but the trial was only held for one day.
Last month, the high court in New Delhi, India, heard testimonials that facial recognition technology the police were using only had 2% accuracy. In India, police are using facial recognition to find missing people.
As facial recognition technology becomes one of the most important tools that police and intelligence agencies have, the industry could see massive growth. The big question remains, though: Who will build and deploy these tools successfully?