Monday, August 2, 2021

A biped robot 5 kilometers run

 On a single battery charge, this biped robot can run 5 kilometers in under an hour.

cassie robot
image: Wikimedia

For the time being, it won't be able to defeat you in a race, but it's a start.

Robotics continues to improve, and while Boston Dynamics' products are usually the most popular, at least on the internet, numerous colleges and universities are working on their own creations. The biped robot at Oregon State University in the United States has already taken a significant stride forward.

Or, to be more precise, a series of steps. Cassie, the university's robot, has successfully completed a 5-kilometer route while running at a modest but constant speed. A single battery charge was used to complete the course, which took 53 minutes.



This is the first bipedal robot to perform such a long outdoor run, according to Agility Robotics, Cassie's developers in college. According to its inventors, the aim for this robot, which is not totally autonomous but operates via remote control and machine learning, is that it will continue to improve until similar robots can deliver mail and assist with home duties.

Biped robots have the advantage of being able to access areas that would be impossible for a quadruped robot to reach, but they are less stable. However, the thought of such robots as office assistants or, in the future, even in homes assisting individuals with mobility disabilities or the elderly is intriguing, and this is precisely what Cassie's developers are striving for in their development.

Cassie won't be particularly quick, at least for the time being, but she could be incredibly handy.

Curated By Gerluxe

AI-based drone defeats human pilots

An AI-based autonomous drone defeats two human pilots, in an astonishing race


Researchers at the University of Zurich (Switzerland) created an algorithm that allowed an artificial intelligence-piloted drone to outperform a human pilot. The drone is a quadrotor, which means it has four rotors, allowing autonomous flying to be faster than that of a human, according to a statement.

Despite the benefits of its development, the drone lacks sufficient autonomy, making its battery one of its weak aspects. This is a regular occurrence with today's drones, which forces them to do duties as quickly as feasible.

This problem, however, does not appear to be as critical as it is with other drones because of an algorithm that quickly leads the drone through an artificial intelligence system. The device can conduct "time-optimal trajectories that completely consider the limits of drones," according to Davide Scaramuzza, head of the university's Robotics and Perception Group.

Philipp Foehn, a PhD student and the study's lead author, stressed that the algorithm "tells the drone to pass all landmarks, but not how or when to do it."

The researchers put their algorithm to the test in a race against two human pilots in the study. Prior to the testing, the three participants were able to practice on the circuit in order to compete on an equal footing.

Finally, the race was held, and the drone piloted by the new algorithm outperformed the humans. Its performance is also more consistent, according to the researchers. This research is the first time in history that an autonomous quadrotor has won a race against human pilots.

Curated By Gerluxe

Artificial Intelligence myths

Artificial Intelligence myths, what is true, and what is a lie?

artificial intelligence
Image: pxfuel

Artificial Intelligence was designed and evolved in response to the growing desire to boost our daily productivity. And, like any other major technology, there are myths and urban legends that need to be debunked, as well as urban legends and other urban legends that turn out to be incorrect. We will attempt to shed some light on this topic in this essay.

Regardless of how the public thinks about AI, it is here to stay, and in fact, as time passes, more and more people are beginning to trust it. This mistrust is partly due to science fiction shows and films that have presented conceptions about AI that are not true, therefore let's look at which of these myths you should dismiss.

The most common misconceptions regarding artificial intelligence

Artificial intelligence is just a machine-based replication of human intelligence. 

One of its goals is for machines to be able to reason rather than just generate logical conclusions and hypotheses from a set of data points.

AI stands for Artificial Intelligence.

Artificial intelligence products are widely available in today's consumer market, and we may discover them in smartphones, PCs, and Internet of Things devices without looking any farther. Industry and astronomy, on the other hand, are a promising area; for example, it played a key role in the 2017 NASA/Google discovery of the planet Kepler-90i, which was the first time this technology was used in astronomy.

However, misconceptions regarding AI have caused considerable alarm over the years, and many people believe that AI's potential is limitless, and that we could end up in a situation similar to the one depicted in the popular Terminator films, with Skynet. Let's see how much of this is a falsehood and how much truth there is in these myths and beliefs.

Humans will be replaced by machines in the workplace.

This is one of the most divisive artificial intelligence myths. It holds the belief that workers and machines are in rivalry, and curiously, this is a myth that seems to be most popular among the younger generation.

According to Fox Business, despite the fact that Amazon's retail warehouses employ more than 200,000 robots as of 2019, the company's rate of employing human employees climbed by 23% between 2019 and 2020, and continues to rise. What is the significance of robotics in this context? One obvious solution to this topic is that businesses are seeking for ways to lower their employees' workloads rather than completely replacing them, which is regarded unachievable.

Indeed, "collaboration" between humans and machines has a long history, with another example being in courier firms, where an automated system classifies parcels by weight and size, with the machines automatically conveying the heaviest packages to relieve humans of the effort. In practice, this is more of an alliance than a replacement, because, as previously said, no AI can completely replace a human being.

Artificial Intelligence (AI) will be the most powerful force on the planet.

When dealing with this myth intellectually, we must be cautious. While several prominent AI specialists have warned of the impending hazards of a machine's ability to think, do we truly comprehend their concerns?

The concept of making machines that are smarter than men and women is divisive. People like Stephen Hawking and Nick Bilton worry that AI will eventually escape our control, escalating fears of a coming machine apocalypse as depicted in numerous sci-fi films.

Elon Musk, the CEO of Tesla, countered with more rational arguments regarding AI regulations and checks and balances. He compared AI to a creature that we could lose control of if we don't handle it properly, but it all relies on how far humans are willing to go in developing something they can't control.

Although AI's strengths are efficiency and accuracy, it is clear that it will never be able to match a person's degree of intuition and emotion, therefore its "mystical (not logical) punch" is determined by what we do with it. Another important point is that AI could fail us, and there has already been solid proof of AI failure in medicine, such as IBM's supercomputer WATSON's unsuccessful cancer therapy prescription.

Artificial Intelligence and the myths of "super humans."

 A female robot character in the 2014 film "The Machine" goes nuts after attaining a high level of emotional intelligence. People's thoughts on the future of AI are shaped by sci-fi movies like these, yet basing our beliefs on our imaginations is illogical.

It is undeniable that AI has influenced decision-making processes in business, astronomy, medicine, and pharmaceuticals, but the fact remains that no matter how well you "train" a machine (because its learning is ultimately dependent on the training it receives), it will never be able to think for itself.

This is an AI constraint that will take many, many years to solve, and many people believe it will never be overcome. As a result, most AI-powered systems will always rely on human judgment to make judgments.

As a result, it is currently ridiculous and impossible to believe that AI's potential can ever create super humans or take over the planet. It makes more sense to represent AI as modern servants that properly and efficiently accomplish duties that would be overwhelming for humans, but are not wiser than us.

The terms artificial intelligence (AI) and machine learning (ML) are interchangeable.

Another misconception concerning Artificial Intelligence is that it is synonymous with Machine Learning; however, the two concepts are not synonymous. Artificial Intelligence and Machine Learning have their roots in the 1950s (the term "machine learning" was invented by Arthur Samuel of IBM in 1952) after the successful development of a chess-playing program.

Chip with Artificial Intelligence

In a nutshell, AI is a machine capable of reasoning in the same way that humans do, and Machine Learning is a subset of AI in which people "train" machines to detect patterns based on data and make predictions, no more and no less.

Artificial Intelligence's sole products are robots.

Again, science fiction films and programs are to blame for this urban legend. The term Artificial Intelligence conjure up pictures of robots in our minds because of how they present this concept to us. AI, on the other hand, is applicable to all areas of technology. Beyond robotics, AI allows for even more complicated creations, such as the facial and fingerprint recognition systems in smartphones, the "smart" devices that many people have in their homes, smart healthcare and business intelligence equipment, and so on.

AI can be used in a variety of ways, including robotics. We can use the term robotics to refer to machines capable of executing physical (and necessary physical) activities automatically in some situations, which is why robotics and AI are sometimes used interchangeably.

In essence, robots do not always have to be AI-based, as they might be made out of mechanical and electrical components (e.g., the robotic arm on an assembly line that simply welds one component to another). AI can only influence how these robots behave when it is used, but it is not required for a robot to have AI.

These are five of the most common artificial intelligence myths today, and we hope that this article has helped to debunk any of your concerns or misgivings about it.

Curated By Gerluxe

A biped robot 5 kilometers run

 On a single battery charge, this biped robot can run 5 kilometers in under an hour. image: Wikimedia For the time being, it won't be ab...