New research, led by the U.S. Army Research Laboratory along with the University of Central Florida Institute for Simulations and Training, is shedding light on the level of trust humans have for robots. The new project focused on the relationship between humans and robots, and whether humans give more value to a robot’s reasoning or its mistakes.
The new research looked into human-agent teaming, or HAT, and how human trust, workload, and perceptions of an agent are influenced by the transparency of those agents such as robots, unmanned vehicles, and software agents. Agent transparency is when a human is able to identify the intent, reasoning process, and future plans of agents.
The new research suggests that human confidence in robots decreases whenever the robot makes a mistake. This is regardless of whether or not the robot has been transparent with its reasoning process.
The new research was published in the August edition of IEEE-Transactions on Human-Machine Systems. The paper was titled “Agent Transparency and Reliability in Human-Robot Interaction: The Influence on User Confidence and Perceived Reliability.”
Traditional research dealing with human-agent teaming uses completely reliable intelligent agents that make no mistakes. However, this new study was one of the few that explored how agent transparency interacts with agent reliability. The study involved a robot that made mistakes while humans were watching, and the humans were then asked if they viewed the robot as less reliable. During the entire process, the humans were given insight into the robot’s reasoning process.
Dr. Julia Wright is the principal investigator for the project, and she is a researcher at U.S. Army Combat Capabilities Development Command’s Army Research Laboratory, or ARL.
“Understanding how the robot’s behavior influences their human teammates is crucial to the development of effective human-robot teams, as well as the design of interfaces and communication methods between team members,” she said “This research contributes to the Army’s Multi-Domain Operations efforts to ensure overmatch in artificial intelligence-enabled capabilities. But it is also interdisciplinary, as its findings will inform the work of psychologists, roboticists, engineers, and system designers who are working toward facilitating better understanding between humans and autonomous agents in the effort to make autonomous teammates rather than simply tools.”
This new project was part of a larger one known as the Autonomous Squad Member (ASM) project that is sponsored by the Office of Secretary of Defense’s Autonomy Research Pilot Initiative. The ASM is an actual small ground robot that is used within an infantry squad. It is able to communicate and interact with the squad.
The study involved participants observing human-agent soldier teams in a simulated environment. The ASM was part of the team, and it moved through a training course. The task for the observers was to monitor the team and evaluate the robot. Throughout the training course, the team was presented with various different events and obstacles. The soldiers were able to navigate each one correctly, but there were times when the robot could not understand the obstacle and made mistakes. The robot then sometimes shared its reasoning behind certain actions as well as the expected outcome.
The study found that the participants were more concerned with the robot’s mistakes compared to the underlying logic and reasoning behind them. The robot’s reliability played a major role in the participant’s trust and perceptions. Whenever the robot made a mistake, the observers rated it’s reliability lower.
The reliability and trust increased whenever the agent transparency was increased, or when the robot shared details and reasoning behind its decision. However, the reliability and trust was still lower than robots that never suffered an error. This suggested that the sharing of reasoning and underlying logic could help with some of the trust and reliability issues surrounding robots.
“Earlier studies suggest that context matters in determining the usefulness of transparency information,” Wright said. “We need to better understand which tasks require more in-depth understanding of the agent’s reasoning, and how to discern what that depth would entail. Future research should explore ways to deliver transparency information based on the tasking requirements.”
This new research will play a critical role in the field because of the increasing interaction that is taking place between humans and robots. One of the areas which will be the most important is the military. As seen in these exercises, robots and soldiers are eventually going to be side by side. Just as a soldier has to have trust in another soldier, the same will apply to robots. If that is able to be achieved and robots became commonplace in infantry squads, it will be another instance of artificial intelligence penetrating the defense industry.
Chadwick Xu, Co-Founder & CEO of Shenzhen Valley Ventures – Interview Series
Chadwick Xu, is the Co-Founder & CEO of Shenzhen Valley Ventures, a hardware-focused company for engineers, by engineers. They partner with startups to prototype, design for manufacturing, test, debug, and manufacture their hardware products.
When did you first want to become an entrepreneur?
From day one when I first came to Shenzhen in 1992, and this is the very reason I came to the city. The path I followed is similar to many other young graduates (at that time), working for a corporate to learn, accumulate expertise, and build connections, then they start their own business.
You founded Zowee Technology in 2004 which was officially listed on the Shenzhen Stock Exchange in 2010. Could you give us some details on what Zowee is and the journey you had creating such a successful company?
The founders of Zowee knew each other by doing business, each partner had their own business before Zowee was founded. The idea is that, instead of having many small businesses, it made more sense to combine the small businesses together and build a bigger company, hence Zowee was born.
Zowee was positioned as a contract manufacturer, and it has remained to be so till today. Besides the effort we put into it, I think the major driving power of its success is the overall economic environment. The 1990s and 2000s was the golden era for Chinese to enter into private businesses. China’s opening up policy and the strong global demand created an ideal opportunity.
Shenzhen Valley Ventures (SVV) is your newest venture. What was the inspiration behind this business?
The success of Zowee represents the opportunity of “Made in China”, but in the future, the big opportunity is not in manufacturing, it’s in technology. What Zowee took advantage of was the labor bonus China had, and what SVV is trying to take advantage of is the engineering bonus China now has today. China is still falling behind on research, but China has the most engineering talent (globally). If SVV can put together a mechanism, enabling the best engineering resources to support the advanced research, so that research could be commercialized faster, it will be a mutually beneficial business structure.
Furthermore, China is becoming the largest single market, both in B2C and B2B business, if SVV’s engineering platform could help international deep-tech startups enter the Chinese market more efficiently, that will be another bonus.
What are some of the more interesting companies that are using SVV’s batch manufacturing services?
There are three groups of projects that have been using SVV’s small-batch manufacturing services:
- Startups: Generally, there is a long growing curve for startup companies to reach the volume that a contract manufacturer is willing to take. SVV has been helping startups when they are still very small and in the early stages of their delivery. This group is the group SVV has been supporting the most. e.g., Neosensory Buzz, a wearable device maker that helps people suffering from hearing or visual problems can now “hear” environmental noises with their skin, and eventually improve their living quality. Sutro, a water quality monitoring device that allows users to access their pool water quality data 24/7, streaming wireless to their cellphone.
- Corporate innovation projects: These projects are identified as “testing the water” projects in a corporate, it may or may not convert into a commercial project, depending on how the testing goes. e.g., an IoT box B/S/H tested to convert non-connected home appliances into connected smart appliances, or the air quality data monitoring device MANN+HUMMEL plans to embed into their existing air filters.
- Universities and institutional research projects: Some researchers need to have physical devices built and deployed to validate or prove the research they made, e.g., an earthquake forecast system/algorithm Beijing university have been researching, SVV helped them to design and build several hundreds of units, these testing units have been deployed nation wide in China, helping researchers collect huge amounts of valuable data for their studies. From 2019, international institutes joined in the study program, and more units are being deployed in Taiwan, Japan, Pakistan, India, and other countries.
SVV enables companies to use their services to prototype products before full scale production launches. Could you elaborate on why SVV is a good partner for prototyping versus other competitors?
SVV is not only a prototyping service provider, instead, it’s a full-cycle development platform provider, including prototyping, engineering, development, testing, certification pre-scanning, pilot production validation, and commercial-ready small-scale production.
This full cycle platform is usually owned by large-scale manufacturers, for the business of large-scale orders from large brands. A innovation dedicated full-cycle platform is a scarce resource, not only for startups, but also for corporations “testing the water” innovation projects.
Can you discuss how SVV is using computer vision to guide manufacturing robots towards better assembly?
SVV is not deploying computer vision guided robotics on its own production lines since SVV’s facilities are used only for small-scale production and for less mature products, the products of such products are more efficient with using manual assembly, since robots are more efficient in producing higher volume and mature designed products. However, SVV has been supporting computer vision guided robotic startups throughout their development.
At a recent Web Summit panel, you discussed how robots will soon be taking over manufacturing from humans, due to how precise robots are. How long until humans are completely removed from the assembly process?
It’s difficult to predict a precise point in time, but it may come much faster than most people think, and it will come gradually, the more standardized the assembly process becomes, the earlier it will be replaced by robots, and gradually will get to the level of fully automated processes.
Just as shown in “American Factory”, there will be a longer period to have robots and humans working together, with the trend of more and more humans replaced by robots, if you stay in a factory for a long period, this replacement process is quite visible.
What are your views on the current AI rivalry between the United States and China?
Competition is always good for the economy and technology, and in almost every competition scenario, the output is not like boxing, with the end result as a lose or win, instead, it often ends up as a symbiotic relationship, each party ends up with its own differentiated advantage.
The past several decades has been the history that China gradually catches up from low-end to middle-end innovation, in areas such as, telecommunications, to directly compete with the US, but the US still takes the leading position in most areas. In the 1800’ and 1900’, US and EU worked very hard, just like the Chinese are doing now, but the wealth and technology advantage somehow created a lay back, welfare dependent society. In many cases, China’s catching up serves as a power to push the US and EU out of their comfort zones and drive them to be more innovative to keep the leading position.
Is there anything else that you would like to share about SVV?
I have been reiterating my opinion on jobs being taken by robots, most discussions are highlighting the bad side, taking jobs from people, increasing the joblessness percentage, creating severe social problems. But on the good side, imagine if smarter machines were built, to improve production efficiency unlimitedly high, constantly growing until the machines can produce an abundance of materials for everyone on this planet, that is supposed to be the bright ending for the AI revolution.
This is almost exactly what Karl Marx describes as the ideal world, “From each according to his ability, to each according to his need”. Or, a perfect utopian society. For the first time in human history we have the opportunity to witness the elimination of hunger entirely from Earth.
We are excited that SVV is a part of this new AI revolution, to help more creative ideas become reality. Drops of water can eventually become flowing rivers, the idea of participating and being witness to this new history is hugely rewarding for us.
Thank you for this great discussion on robotics. To learn more readers may visit Shenzhen Valley Ventures.
AI Makes it Easier for Drones to Scan and Excavate Terrain
Researchers from Aarhus University (AU) and the Technical University of Denmark (DTU) have collaborated on a project that aims to decrease the costs of measuring and documenting gravel and limestone quarries, while at the same time being faster and easier than the traditional method.
The project included the use of artificial intelligence (AI), which took over the traditionally human-controlled drones that are currently relied on to complete the task.
Erdal Kayacan is an associate professor and expert in artificial intelligence and drones at the Department of Engineering at Aarhus University.
“We’ve made the entire process completely automatic. We tell the drone where to start, and the width of the wall or rock face we want to photograph, and then it flies zig-zag all the way along and lands automatically,” says Kayacan.
Limitations of Human-Controlled Drones
The current method of measuring and documenting gravel and limestone quarries, cliff faces, and other natural and human-made formations relies on drones to photograph the area. A computer then receives the recordings and automatically converts everything and creates a 3D terrain model.
One of the downsides of this method is that drone pilots cost a lot, and the measurements are time-consuming. In an excavation, the drone pilot has to make sure that the drone keeps a constant distance from the wall. At the same time, the drone camera has to be kept perpendicular to the wall, making it a complex and difficult task.
In order for the computer to convert and create a 3D figure out of the images, there has to be a specific overlap in the images. This is the main process that was automated by artificial intelligence, and it drastically reduced the complexity of completing the task.
“Our algorithm ensures that the drone always keeps the same distance to the wall and that the camera constantly repositions itself perpendicular to the wall. At the same time, our algorithm predicts the wind forces acting on the drone body,” says Kayacan.
AI Overcomes Wind Problem
The artificial intelligence also helps overcome the wind, which is one of the biggest challenges with autonomous drone flight.
Mohit Mehndiratta is a visiting Ph.D. student in the Department of Engineering at Aarhus University.
“The designed Gaussian process model also predicts the wind to be encountered in the near future. This implies that the drone can get ready and take the corrective actions beforehand,” says Mehndiratta.
When a human-controlled drone is completing this task, even a light breeze can alter the course of it. With the new technology, wind gusts and the overall wind speed can be accounted for.
“The drone doesn’t actually measure the wind, it estimates the wind on the basis of input it receives as it moves. This means that the drone responds to the force of the wind, just like when we human beings correct our movements when we are exposed to a strong wind,” says Kayacan.
The research was completed in collaboration with the Danish Hydrocarbon Research and Technology Centre at DTU, and the results of the project will be presented in May 2020 at the European Control Conference.
Researchers Create Soft Robot Able to Change Shape and Roam
One of the challenges surrounding soft robotics is that most of them are required to be attached to an air compressor or plugged into a wall. Researchers from Stanford set out to overcome this challenge.
Nathan Usevitch is a graduate student in mechanical engineering at Stanford.
“A significant limitation of most soft robots is that they have to be attached to a bulky air compressor or plugged into a wall, which prevents them from moving,” said Usevitch. “So, we wondered: What if we kept the same amount of air within the robot all the time?”
The team was able to develop a human-scale soft robot that is capable of changing its shape. By doing this, the soft robot can latch onto and handle objects, and it is able to roll in controllable directions.
The research was published in Science Robotics on March 18.
“The casual description of this robot that I give to people is Baymax from the movie Big Hero 6 mixed with Transformers. In other words, a soft, human-safe robot mixed with robots that can dramatically change their shape,” said Usevitch.
This soft robot was developed by combining three different types of robots. The simple version of the team’s invention is called an “isoperimetric robot,” since the shape changes while the total length of the edges and the amount of air inside stays the same.
The isoperimetric robot was developed out of soft robots, truss robots, and collective robots. Each category of robotics brought a different advantage: soft robots are lightweight and compliant, truss robots can change shape, and collective robots are small and collaborate.
Sean Follmer is an assistant professor of mechanical engineering and co-senior author of the paper.
“We’re basically manipulating a soft structure with traditional motors,” said Follmer. “It makes for a really interesting class of robots that combines many of the benefits of soft robots with all of the knowledge we have about more classic robots.”
The team also developed a more complex version of the robot by attaching several triangles together. They were able to coordinate the movements of the different motors, which allowed the robot to carry out desired behaviors, such as picking up a ball.
Elliot Hawkes is an assistant professor of mechanical engineering at the University of California, Santa Barabara and co-senior author of the paper.
“A key understanding we developed was that to create motion with a large, soft pneumatic robot, you don’t actually need to pump air in and out,” said Hawkes. “You can use the air you already have and just move it around with these simple motors; this method is more efficient and lets our robot move much more quickly.”
According to Zachary Hammond, a graduate student in mechanical engineering at Stanford and co-lead author of the paper, one of the possible uses for this soft robot is space exploration.
“This robot could be really useful for space exploration — especially because it can be transported in a small package and then operates untethered after it inflates,” said Hammond. “On another planet, it could use its shape-changing ability to traverse complicated environments, squeezing through tight spaces and spreading over obstacles.”
The researchers are now trying out different shapes, and they want to test the robot in water.
Allison Okamura is a professor of mechanical engineering and co-author of the paper.
“This research highlights the power of thinking about how to design and build robots in new ways,” said Okamura. “The creativity of robot design is expanding with this type of system and that’s something we’d really like to encourage in the robotics field.”
- Shell Begins to Reskill Workers in Artificial Intelligence
- Anthony Macciola, Chief Innovation Officer at ABBYY – Interview Series
- Brain Implants and AI Model Used To Translate Thought Into Text
- Marc Sloan, Co-Founder & CEO of Scout – Interview Series
- AI Developed to Translate Brain Activity into Words