Autonomously identify elevators going up and down, deliver materials to the door of designated households, call households to pick up goods after arrival, and return to the lobby on the first floor to charge independently… This is the daily work scene of the epidemic prevention and distribution robot in the Zijunfu Community, Baihedong Street, Liwan District, Guangzhou .

The workload of material distribution in quarantined communities is huge. Recently, videos of 13 anti-epidemic distribution robots on duty have spread all over the Internet. “The intelligentization of the epidemic area is online, this is very Chinese!” The owners praised.

Science and technology help fight the epidemic. The Qianxi Robot Group, a subsidiary of Country Garden, completed the delivery task of “opening up the last 100 meters” in just 72 hours, and it is expected to serve other isolated communities in the future. Behind the “hard-core” fight against the epidemic is the responsibility of Country Garden to work together to fight the epidemic.

On the one hand, actively play the leading role of community party members, mobilize the red forces to participate in epidemic prevention work, and play the role of grassroots party organizations as a battle fortress; on the other hand, in this smoke-free war, care is done in detail, showing warmth everywhere.

13 “broad beans” epidemic prevention and distribution robots are on duty

On the eve of the Dragon Boat Festival, the owner group and circle of friends of residents in the isolated community in Liwan District, Guangzhou suddenly became lively, and 13 delivery robots “broad beans” were officially put on duty in the Zijunfu community.

The owner, Mr. Wang, received an unfamiliar call that day, and a sweet voice reminded him to open the door to pick up supplies. “I opened the door and looked at it. I was surprised. What I saw was not a fully armed property manager, but a cute robot.” Then, he saw the robot swiping message in the owner group.

This distribution robot named “Fava Bean” has multiple functions such as intelligent map navigation and obstacle avoidance, elevator ride, errand delivery, etc. It can efficiently serve the three distribution needs at one time, so that residents in closed management can achieve it at their doorstep. No-contact food fetching throughout the entire process.

It is understood that there are 9 buildings, 13 units, and about 1,000 owners in Zijunfu Community. Each building is delivered at least 7-8 times a day. Because each material arrives at a different time and quantity, sometimes a household’s materials have to be separated. Multiple delivery.

“Since the quarantine in late May, every property housekeeper has worked overtime without rest. The work pressure is very heavy. The arrival of the delivery robot has not only greatly reduced our workload, but also increased our security.” Community Property Management Staff introduction.

“After completing the deployment of the Zijunfu community, we will also provide services in other isolated communities in accordance with the requirements of the government’s epidemic prevention department.” said Xiao Ran, secretary of the party branch and deputy general manager of the Qianxi Group under Country Garden.

Entering the community for the first time for distribution, effectively overcoming technical difficulties

In February, when the epidemic prevention and control was the most intense last year, the demand for food and beverage supply in parts of Hubei was urgent. Qianxi Group and Guoqiang Public Welfare Foundation donated two batches of self-developed container claypot rice robots to Hubei as soon as possible. Through fully automatic and contactless 24-hour hot meal supply, they provided a large number of medical staff such as Guangdong Medical Team to Hubei. Provide catering services to help Hubei win the battle against the epidemic.

The war “epidemic” continues. On the night of June 4, Qianxi Group, a subsidiary of Country Garden, learned from the epidemic prevention department of the Guangzhou Municipal Government that the distribution of materials in quarantined communities is huge and urgently needs to solve the problem of the final “100-meter” distribution in quarantined communities.

With the strong support of the government’s epidemic prevention department and the isolation community, Qianxi Group basically completed the overall deployment of robot distribution in just 72 hours. After two days of intense debugging and operation, the project was delivered in the early morning of June 10 and the robot was delivered. Also officially on duty.

Xiao Ran, Secretary of the Party Branch and Deputy General Manager of Qianxi Group, introduced, “Since April this year, this delivery robot has been stationed in many hotels in Panyu and Huadu, Guangzhou, responsible for 24-hour delivery tasks and providing services for overseas inbound quarantine personnel. This is the first time that it has entered the community to carry out material distribution services.”

In the early morning of June 8, the R&D team was intensively debugging

There is a big difference between the application scenarios of the community and the hotel. For example, in a hotel, the robot can automatically dial the fixed phone to remind the room after reaching the door of the room. However, the communication methods of the residents in the community are very different, which increases the technical difficulty. Qianxi Group invested 3 R&D personnel in the Within 4 hours, the problem of adapting household mobile phones was overcome.

In addition, the on-site high-density residential elevator control signal and the robot equipment signal interfered with each other, which also brought certain challenges to equipment deployment. Two R&D personnel provided solutions and test plans within 5 hours.

Property management, the party and the masses, strengthen the community’s epidemic prevention line

Starting in late May, in accordance with the epidemic prevention and control requirements of governments at all levels, relevant areas in Guangzhou and Foshan have successively carried out nucleic acid testing for all employees in an orderly manner. Country Garden attaches great importance to the prevention and control of the epidemic, actively plays the leading role of community party members, mobilizes the red forces to participate in the epidemic prevention work, and plays the role of a fighting fortress of the grassroots party organization.

At 5 a.m. on May 29th, Guangzhou Country Garden on Nanpu Island received the notification of nucleic acid collection, and then acted quickly. At 6 a.m., personnel from various departments were called to quickly implement prevention and control measures and carry out a clear division of labor. The Phoenix Butler informs the prevention and control situation in the group, reduces the anxiety and anxiety of the owners after learning the news of the blockade, and soothes the owners’ emotions; the property personnel and volunteers are fully assembled on duty to expeditiously set up nucleic acid detection points in the community. At 10 o’clock in the morning, the construction of nucleic acid detection points in the community was completed, and the intense nucleic acid collection work was carried out in an all-round way. As of 17:00 on the 30th, the nucleic acid collection work in Guangzhou Country Garden Community was successfully completed.

In just two days, 5640 homeowners, 11 gardens, Guangzhou Country Garden Community completed the nucleic acid collection work efficiently, demonstrated the “Country Garden Speed” of community epidemic prevention, and deeply “circled fans” owners.

Huang Yansong, General Party Branch of Guangzhou Country Garden Community, Guangzhou Country Garden Property, stated that the nucleic acid testing operation in Guangzhou Country Garden Community was jointly carried out by the Guangzhou Country Garden Community Party General Branch and the Party Branch of Guangzhou Country Garden Branch. During this period, a large number of owner volunteers joined the voluntary service operation. China, assisting the nationwide nucleic acid collection work.

On May 29th, Nanhai District of Foshan carried out a large-scale nucleic acid test on the permanent population of Guicheng Street. The Party branch of the Yongsheng Project in the Foshan District of Country Garden learned that the Xiabei community to which the project belongs needs to carry out nucleic acid testing, and the public servants and community volunteers stationed in the village have heavy tasks and hope to receive support.

The party branch of the Yongsheng Project immediately responded to the call of the community, and combined with Country Garden Services to set up a volunteer service team to go to the community detection point to support. Due to the heavy inspection task and the large flow of people, in order to better maintain the on-site queuing order, they helped the people with disabilities, the elderly and pregnant women in the team arrange green channels, assisted in answering the doubts of the residents, and helped the elderly to register information, etc., “Red Vest We” have always been on the front line of fighting the epidemic.

In the process of promoting vaccination service activities, the community also played a leading role in party building. From May 29 to 31, Country Garden Phoenix City assisted neighborhood committees to carry out vaccination activities for community residents, and a total of 9,500 people were vaccinated. The property center mobilized 50 people to cooperate vigorously with the government to help the community successfully complete the vaccination activities.

Keep watch and help, fight the epidemic with strength and temperature

While science and technology are helping to fight the epidemic, in various communities in the Guangqing/Dongguan-Shen region, Country Garden attaches great importance to the prevention and control of the epidemic. In this smoke-free war, the care is done in detail and showing warmth everywhere.

At 4 pm on June 2nd, a sudden storm swept across. At this time, Guangzhou Country Garden was carrying out the second round of nucleic acid testing, and the 63-year-old Uncle Wen was responsible for the garbage cleaning at the testing site.

Suddenly in heavy rain, Uncle Wen hurriedly draped the black garbage bag as a “raincoat” on his body. Although he was wearing a hat on his head, he still could not stop the “attack” of the violent storm, and soon his whole body was soaked. Despite this, Uncle Wen did not leave his job.

A kind proprietor saw this scene and came over and gave him an extra umbrella. However, when he turned around and saw the passing owners getting caught in the rain, Uncle Wen didn’t hesitate to give the umbrella to the other party.

In this “epidemic” of the nationwide war of mutual assistance, touching scenes are taking place. The major communities of Country Garden race against the epidemic against the clock, try their best to block the chain of transmission, and work together to fight the epidemic.

In order to reduce the risk of crowd gathering and reduce the waiting time for owners to wait in line, Country Garden Property conducts time-based inspection arrangements for owners according to the garden area.

The Phoenix Butler, through the Moments of Friends and the group of owners, will issue notifications of inspection arrangements around the clock. Volunteers and property management security will go to each building in each garden area from time to time to remind the next batch of owners to prepare to go to the inspection point.

For owners with limited mobility, in addition to opening a green channel for young and old, Country Garden’s property management team also teamed up with medical staff to conduct nucleic acid tests door to door.

During the high temperature in summer and concentrated rainfall in the “dragon boat water”, during the inspection period, Country Garden property personnel urgently “fought” and arranged for personnel to carry sandbags to stabilize the tent.

From early morning to night, the property housekeeper fought side by side with the medical workers to overcome the difficulties together with the community workers. In this special period and at a critical juncture, they built a strong wall for the community owners to prevent the epidemic.

In order to effectively assist the epidemic prevention and control work in Guangdong Province, and at the same time convey the warm care of the Dragon Boat Festival and demonstrate corporate responsibility, on June 11, Country Garden Group and Guoqiang Public Welfare Foundation cooperated with Country Garden Jiangzhong District, Guangqing District and Bixiang Agricultural Company The community initiated anti-epidemic actions and other units, and related units in Guangzhou and Foshan donated anti-epidemic materials to provide guarantee for the stability of logistics.

As the Guangzhou-Foshan epidemic slows down, in this race against the virus, we have seen the “united efforts” of the people of Guangzhou, as well as the efficient action and social responsibility of responsible companies.

Southern Reporter Wang Yanling

Some people questioned “How does this look like a catwalk? It’s just like other robot dogs walking!” In fact, the seemingly ordinary steps are actually hidden mysteries. It does not use a complex algorithm structure such as Boston Dynamics, but reproduces the reflex behavior of cats when they walk, and uses the characteristics of animal nerves and muscles to gradually “cat-like” robot dogs. Not only that, but also published a paper on “Frontiers in Neurorobotics”. Re-engraving the cat, walk out the pace that six relatives do not recognize If you want to take a catwalk, you must first know how a cat walks. When the cat’s jiojio steps on the ground, the knee muscles are stressed, and the waist will move backward; if the waist muscles are stressed, the knees will stretch, and the two parts will be automatically linked. Woke up. The loop that extends from the knee to the waist, and in turn, the loop that extends from the waist to the knee. This crossover function is called the “reflex circuit” in research, and it can produce a stable gait and leg movement trajectory. Recent studies have also shown that cats do not need complex interactions of nerves within the brain when walking, but instead use this reflex circuit to perform stable movements. Reflex circuit and reproduction of muscle characteristics Then the problem comes again. The robot dog has no muscles or nerves. How to apply this “reflex circuit” to the robot dog? There are two main problems to be solved here. First, the modeling of muscle characteristics. The Japanese brother used the Hill muscle model to reproduce the muscle characteristics of the robot dog. In the model, the two legs of the robot dog reproduce the linear joints with virtual rotating hip joints and hips and feet. Extensors and flexors are placed on each virtual joint. The muscles of the virtual rotating joints and linear joints are respectively. Corresponds to the cat’s waist muscles and knee muscles. After recreating the muscle model, it is necessary to establish a “reflex circuit” model between the robotic dog’s waist and knee extensors. The Japanese brother has embedded the excitatory reflex pathway in the muscle model. If the knee joint receiver senses the ground reaction force, the muscle will extend the signal to the waist receiver; conversely, if the waist extensor muscles are fully extended and produce tension, Then the muscles will “feel” the signal to start, and transmit it to the knee joints to change the posture. Body: Font size 16, dark gray (R52 G52 B52), aligned on both sides, margin 8, line spacing 2 Robot dog’s “catwalk” walking experiment Come on, the models are all built, and start walking on the treadmill! Put the robot dog on the treadmill without pressing the switch. Once the receiver on the foot obtains the reaction force from the ground, it immediately activates the reflection circuit. The first two steps were a bit staggering, but the more you go, the more stable you go. The robot dog is equipped with a microcontroller (Teensy 3.6) to provide advanced control. The angle information received from the motor driver simulates the virtual muscle and reflection circuit, and commands the target torque to reproduce its characteristics to the motor driver. When the reflex circuit of the robot dog is cut off, the dog’s walking will become unstable, and the phenomenon that the feet cannot be lifted from the ground occurs. The Japanese brother who makes the robot dog walk catwalk is a graduate student of the Department of Mechanical Engineering of Osaka University. He said: “Most of the four-legged robots use the central transmitter (CPG) for stable walking. Our research proves that it is novel. The reflection circuit can also give the “robotic dog” a stable gait, realizing a robot that can move as efficiently and flexibly as an animal.” [Reference link] https://www.frontiersin.org/articles/10.3389/fnbot.2021.636864/full https://www.youtube.com/watch?v=-iLHRhvDccA This article is reproduced with permission from Robot lecture hall (ID: RoboSpeak ), the original title is “The robot dog gradually “cats”,” using simple reflex circuits and muscle reproduction to make the robot dog walk catwalks! 》

The disinfection robot combines two methods of UVC and chemical disinfection to help kill bacteria and reduce the spread of disease. The robot also has the ability to move flexibly, remotely or pre-programmed for a certain sterilization route. Le Anh Son, Deputy Director of Phenikaa Institute of Research and Technology (Phenikaa University), said that the robot is also capable of taking on tasks to support epidemic prevention and control in Bac Giang such as: Transporting necessities , monitoring sick and isolated people through monitoring cameras. Phenikaa University’s disinfection robot. The robot is fully self-propelled, avoiding collisions by using LiDAR technology, integrating artificial intelligence to recognize and remind people to wear masks, measure body temperature and a host of other superior features. . On the evening of June 2, the robot was put into operation for the first time. Before that, in the middle of May 2021 when the epidemic was complicated, students of Phenikaa University launched the campaign “Online learning – Show talent against the epidemic”. The students spent time and effort to create a droplet shield for medical staff, medical staff at hospitals, treatment facilities, and concentrated isolation. Up to this point, 10,000 droplet shields have been sent to more than 30 localities, of which more than half have been transferred to the epidemic center Bac Ninh, Bac Giang. High-tech products “make in Vietnam” such as: COVID-19 epidemiological information maps, disinfection robots, self-propelled robots, smart school solutions … or simply splash shields, orezol water These are all feelings of love and sharing from the collective staff, lecturers, scientists and students of Phenikaa University sent to the epidemic center.

Doctors at the No. 2 field hospital study how to operate the Phenikaa-X disinfection robot. Photo: C.Le Dr. Le Anh Son, Deputy Director of Phenikaa Institute of Research and Technology, said that the disinfection robot can take on the task of supporting epidemic prevention and control in Bac Giang with the main tasks such as: Transporting necessities, monitor patients and quarantined people through monitoring cameras. The robot is fully autonomous and avoids collisions by using LiDAR technology, with built-in artificial intelligence to recognize and remind people to wear masks, measure body temperature and a host of other features. Doctors at the No. 2 field hospital study how to operate the Phenikaa-X disinfection robot. Photo: C.Le Disinfection robot is one of the high-tech products and is directly applied in the prevention and control of Covid-19 epidemic by scientists from Phenikaa Institute of Research and Technology, Phenikaa-X Joint Stock Company, Phenikaa University research and develop to apply in real life. Doctors of field hospital No. 2 said that the presence of the Phenikaa-X disinfection robot helps the hospital a lot because it has an extra protective shield. This robot is run continuously in the corridor to help disinfect the air. Doctors at the No. 2 field hospital study how to operate the Phenikaa-X disinfection robot – Photo: C.Le Previously, students of Phenikaa University also made droplet shields for medical staff, medical staff at hospitals, treatment facilities, and concentrated isolation. Up to this point, 10,000 droplet shields have been sent to more than 30 localities, of which more than half have been transferred to the epidemic center Bac Ninh, Bac Giang. In addition, Phenikaa University also donated 10,000 bottles of Orezol water to the frontline medical staff in the fight against the epidemic.

The first robot tractor in operation on the farm in the UK. Photo: TL. According to Fram Famers (UK), commercial robot tractors have been put into operation at Home Farm Nacton (member of Fram), in Suffolk (United Kingdom), from mid-April. Technological improvement required, along with difficulties in finding seasonal labor, led to the investment of robot tractors at the aforementioned farm. Home Farm, which specializes in organic agricultural production, received the Robotti 150D from Agrointelli, a Danish company. It was the first robot tractor to operate in the UK, and is now fully operational on the farm. The Robotti 150D is used to till the soil to prevent weeds from sprouting by using a harrow attachment. In the near future, Robotti 150D at Home Farm will be used to perform many other jobs such as mowing the lawn, punching holes to plant onions, etc. Home Farm produces both organic and conventional vegetables, including leeks, bok choy, onions, beets, potatoes, sugar beets and grains, on 1,940 hectares. The farm is looking to apply more advanced technology to production and business activities. Andrew Williams, Home Farm manager, said: “A large portion of our produce is organic, so we are increasingly limited in how we can control weeds. Organic production requires a lot of repetitive work as the beds are constantly plowed and weeded. Meanwhile, sourcing seasonal workers, mainly foreigners, has become increasingly difficult. We want to look to the future of farming and Robotti fits our setup perfectly.” Designed and manufactured by Agrointelli, the Robotti 150D is controlled automatically by GPS via an integrated computer, independent of the operator, instead following a pre-programmed route in the field. Tractor Robotti 150D. Source: Agrointelli. “Accuracy lies in the GPS mapped field, with Robotti always taking the exact same lines,” explains Frederik Rom, sales manager for Agrointelli. It is equipped with an RTK GPS, so setting up on any farm is simple. It takes about 10 minutes to map the 4-acre field at Home Farm and another 10 minutes to log the weeding plan into the system. It is important that this is correct from the start, as it will determine reliable performance.”

Owning a small house in the middle of a green hill, separate from the noisy city life, is the desire of many people? With this in mind, master craftsman Jeffery Broadfield and a construction team from Smith and Primmmer created a robotic recycling house in the middle of a Berry field (Australia). The house has an area of ​​​​9m2, allowing two people to live together and still have their own space. The structure of the house is made mainly of recycled iron shells taken from a disused wharf, this material is intended to withstand the scorching rays of the Australian sun. The ground floor is a place to cook, store things and have a heating system for the house. The bedroom is located above and is accessed by a ladder. The bathroom is a deck behind the house with its own rainwater tank and toilet with a composting system. Water is collected on the roofs and then directed into a cistern above the bathroom, supplying water to the shower and kitchen. A wood-burning stove in a pot heats the cabin at night. Solar panels on the roof provide energy for lighting. In addition, the house has a ladder leading to the roof and acts as a lightning wire. If viewed from afar, the house looks like a wooden robot in the middle of a green field, realizing the dream of a peaceful resort but still fully equipped for modern people.

Nathan Copeland controls a robotic arm using electrodes implanted in his brain. Published in the journal Science, the team say their work demonstrates that adding sensation to the technology significantly improves the function of prostheses for quadriplegics, compared with based solely on visual cues. Nathan Copeland, 34, told AFP: “I am the first person in the world to have a device implanted in the sensory cortex that scientists can use to stimulate my brain directly. And then, I felt like I had a real feel in my flesh and blood hand.” In 2004, Copeland was involved in a car accident that left him with a severe spinal cord injury and quadriplegia. He volunteered for scientific research and six years ago he underwent major surgery to implant tiny electrodes in his brain. Two sets of 88 electrodes the width of a human hair are arranged like tiny combs and penetrate deep into the cortex, which controls motor function. According to Associate Professor Rob Gaunt of the Department of Physical Medicine and Rehabilitation at the University of Pittsburgh, co-leader of the study, fewer than 30 people in the world have received such a transplant. What’s unique about Copeland’s case is that an additional set of electrodes is connected to his dorsal cortex, which receives and processes sensations. The idea of ​​sending haptic feedback to the tactile sensory cortex dates back decades, but doing so in a controlled and understandable way by brain circuitry used to be a major challenge. After Copeland underwent surgery to install the electrodes, the team was truly thrilled. Then came the decisive moment, when they tried sending the first tactile signal. “It felt really fuzzy,” Copeland recalls. He asked them to try again to make sure it was real. Before the interface could work with the robotic arm, the scientists had to perform a series of tests on Copeland. First, they needed to learn which electrodes cause what sensation when activated and which fingers are associated with them in order to set up the robotic hand correctly. They also made him watch a video of the robotic arm moving left or right and recorded the electrodes lighting up when he was asked to “think” it was him controlling it. Copeland sat next to a black metal robotic arm and was asked to pick up a series of small objects such as rocks and spheres and place them in a box – when the tactile sensors were turned on or off. He can complete each task in an average of twice as fast when the sensors are activated, and can even perform more complex tasks like picking up a glass and pouring its contents into a glass. another cup. The team wanted to further refine the prosthetics because they didn’t want to just do science experiments in the lab, but wanted to actually make devices that would be useful to people. Copeland has set up his brain-computer interface at home when the Covid-19 pandemic shut down universities and has used his downtime to learn how to draw on a tablet and even play video games. death. He does this by using his mind to send signals directly to the computer, instead of using his arms to press buttons.

iRobot S9+. S9+ is the most advanced robot vacuum cleaner from the American brand of robot vacuum cleaners – iRobot. The product comes with a dock that automatically collects garbage in the body. Users therefore only need to empty trash and dust from the dock instead of having to clean the drawer on the robot as often as with conventional products. The product uses the AllergenLock technology trash bag designed specifically for the S9+ to filter allergens such as pollen and mold. Half square and half circle machine, dual omnidirectional brush set, technology that is exclusive to iRobot, recently appeared on Ecovacs T9 AIVI series. Most of the latest technology on robot vacuums is available on the S9+, such as redrawing diagrams with a laser sensor using the vSLAM algorithm; identify obstacles, objects that cause obstacles, entanglement in the course of running; control via smartphone and map selection. Ecovacs Deebot T9 AIVI (13.5 million VND) T9 AIVI is Ecovac’s first robot line to integrate the Aeroforce dual-action rubber brush system shared from iRobot. As a result, the product has a higher suction efficiency, even with large garbage objects or stains sticking to the floor surface. This is also the top model with high suction power today, 3,000 PA, 300 ml garbage box capacity, 240 ml water box for cleaning, 5,200 mAh battery capacity. Compared with high-end robot lines, T9 AIVI has the most significant advantage of a vibrating cleaning system and an electric water regulating valve. The software on the machine also allows storing many different maps, suitable for multi-storey houses. Roborock S7 (12 million VND) Roborock S7. S6 Max proved “short of breath” compared to products from Ecovacs, Roborock has launched S7 with many technologies no less than the competition. The most significant upgrade is the Sonic Mopping water damper vibration mopping system that can vibrate 3,000 times per minute. The new wiping mechanism is significantly more effective than the old S6 Max series. However, the suction power of the machine is only kept at 2,500 PA and there is no dual rubber brush system like iRobot or Ecovacs. The S7 has a camera system that detects objects such as wires, socks, and baby items so that the device does not get entangled or entangled when operating. This camera can also be viewed directly from the phone. Neato Bootvac D7 (19 million VND) Neato Bootvac D7. Bootvac D7 from Neato has many similarities with iRobot S9+ in terms of half square, half round design. The company explains that this design helps to limit the omission of dirt in the corner. This is also a recognizable difference with most of the remaining robot models on the capital market, which are circular in shape. The machine also has a navigation system, laser map redraw, save multiple maps and select a cleaning area. Largest trash can capacity in the segment – 700 ml. However, the suction power of the machine is poor compared to competitors at 1,600 PA. Without the dual rubber brooms like Ecovacs or iRobot, the D7 has a broom that’s almost twice as long. The machine can also be controlled remotely via the application, schedule cleaning, create virtual walls …

Industrial design students Minwook Paeng have just come up with a solution to make it easier to use your smartphone while on the road: the Third Eye robot eye. This is an eye-shaped robot that is fixed on the user’s forehead and detects obstacles in front of them when they use their smartphone on the road. This humorous device allows users to navigate while texting or surfing Instagram without worrying about being obstructed by real-world obstacles. This artificial eye has a translucent plastic shell, with internal components based on the open source electronic platform Arduino. It will be applied to your forehead with a thin gel pad. The device is integrated with a speaker and an accelerometer, which is a sensor commonly found on smartphones used to detect the orientation of the phone and automatically rotate the screen when necessary. The device is integrated with a speaker and an accelerometer, which is a sensor commonly found on smartphones used to detect the orientation of the phone and automatically rotate the screen when necessary. The accelerometer on the Third Eye detects when the user’s head is bent to open the plastic eyelid, revealing a sonar sensor. When detecting an obstacle ahead, the eye will emit a buzzing sound to signal the user. The use of mobile phones while crossing the street is increasing and affects public safety on the street. Even the city of Montclair in Southern California (USA) has introduced a law banning pedestrians from using phones or wearing headphones while walking across the street. First-time offenders will be warned, if they repeat the offense for the second time, they will be fined $100 (about 2.2 million VND), if they repeat the offense many times, the fine will be up to $500 (about 11 million VND). A scientific study has confirmed: people who use mobile phones will walk slower and even go sloppy. Due to paying attention to the phone screen so as not to type the wrong content or read the message correctly, the human observation ability will be narrowed, limiting the ability to detect obstacles and not avoid timely. As a result, accidents can happen whenever, for example, hitting other people or crashing into traffic. Please watch the video: Robot make in Vietnam knows how to do math. Source: VTV

After Charlton ordered the human staff, the second robot delivered the finished food to the customer. The food is neatly arranged on the tray. “Thank you,” Vicki Charlton said. The robot replied “Nothing”. Restaurant Mr. Q in South Florida (USA) uses robots to serve customers. Photo: Sun-Sentinel. For the past several weeks, Joy Wang, the owner of Mr. Q has difficulty recruiting waiters on weekends, when the number of visitors is too large. “A former employee once told me that they would rather stay home and get unemployment benefits,” Wang said. Due to not recruiting enough staff, Wang decided to use robots to assist in serving the restaurant. Last week, she bought three service robots for $30,000. According to the Sun-Sentinel , Mr. Q is one of the first restaurants in South Florida to use robots to serve customers. This trend also extends to Miami. In late March, REEF, an online restaurant service company, partnered with startup Cartken to produce a local food delivery robot. Matt Lindenberger, chief technology officer of REEF, said the electric robot can deliver goods in the downtown Miami area in 30 minutes. According to Lindenberger, the robots are the size of hot dog carts, which can climb up or down sidewalks, automatically avoiding cars and pedestrians. In addition to the reception robot, Mr. Q also equips a robot to serve food. Photo: Sun-Sentinel. “Applying service robots helps to save staff costs. Instead of using 5 people to deliver 5 orders, I can use 3 robots and a human,” Lindenberger said. In California, fast-food chain Chik-Fil-A also plans to use semi-autonomous robots for delivery, while Domino’s will deliver pizza by robot car in Houston. An employee at Mr. Q thinks it’s reasonable to apply robots because they don’t carry disease-causing viruses, don’t cough on food, don’t rest (unless they need to recharge). Instead, they are always happy, only interested in work. Michael Salcedo, waiter at Mr. Q said the older guests enjoyed Peanut. It also takes care of taking orders if another employee is busy. When bringing guests to the table, Peanut’s touch screen will display the menu. On special occasions, the robot can even sing “Happy Birthday” or “Merry Christmas” in four languages. Older diners enjoy being served by robots. Photo: Sun-Sentinel. Unlike Peanut, the restaurant’s two remaining service robots are unnamed. However, Salcedo prefers to call them Beavis and Butthead, because they often bump into each other in restaurants. Shaheen Maleki, restaurant manager Mr. Q said the robots are sometimes quite noisy. “There was a time when Peanut was about to run out of battery. It says ‘I have to go back. I had to go back by myself’, while we couldn’t find the volume button (to adjust the robot’s speakers)”, Maleki said that the service robot both attracts customers and helps the restaurant, but can also be annoying at times. . According to Maleki, many customers film the robot and then post it on the Internet to help the restaurant become more known. With large dining tables, a robot can serve food in one go, the job sometimes requires 3-4 people, or one person has to move many times. Various types of unique robots at the technology fair in Thailand Amazing robotic assistants helping with directions and serving food and medical robots were showcased at a recent tech fair in Thailand.

The delegates cut the ribbon to launch the training program of artificial intelligence and robot technology. (Photo: Thu Hoai / VNA) On April 24, Ho Chi Minh City National University held the launching ceremony of the Training Program artificial intelligence and Robot technology (AIC). The program is invested more than 32 billion VND, is part of the 320 billion VND sponsorship package of Inter-Pacific Group (IPPG) for National University of Ho Chi Minh City and Dalat University to establish a training school. CEO creation and AI education development. Held at the Software Technology Zone of Ho Chi Minh City National University, the STEM, STEAM interdisciplinary curriculum on artificial intelligence (AI) for students in age, level of education, right from grade 1. Regarding the document, the AI ​​Future Intelligent Manufacture syllabus of UBtech education USA Company copyrighted by IPPG, adjusted by the National University of Ho Chi Minh City, Vietnameseized, evaluated and published, put into teaching. This is also a program that has been taught by many advanced countries in the world. The program includes activities such as: AI-Robotics program and summer courses for students; teacher training course AI-Robotics from basic to advanced; organize seminars on AI-Robotics for the community and businesses. This is also a space to experience, introduce and display IT application solutions, artificial intelligence of domestic and foreign corporations, enterprises, and AI scientific products of students. University. Mr. Truong Minh Huy Vu, Director of Software Technology Zone, Ho Chi Minh City National University , sharing the process of Vietnameseization and document appraisal, in addition to the participation of teachers, scientists in the field of computer science and AI, there are also experts in social fields. language, pedagogy so that knowledge is closer to students and students. Building learning materials is a continuous process, so in the coming time, the curriculum of the program will continue to be improved and adjusted to suit with reality. Associate Professor, Dr. Vu Hai Quan, Director of Ho Chi Minh City National University, shared that the unit is currently participating in training and research on artificial intelligence both in depth and in width. In which, in depth, the National University of Ho Chi Minh City is preparing to establish the artificial intelligence research institute; at the same time undertaking the task of training high-quality human resources at international level, in which chaired many component projects on information technology, artificial intelligence … Along with that, the National University of Ho Chi Minh City also focuses on programs in width such as implementing training programs for artificial intelligence and technology. Robot , to spread knowledge about AI more to the community, especially among students; thereby provoking, transmitting passion for scientific research to students, discovering and fostering new factors in the field of AI research. Ms. Le Hong Thuy Tien, General Director of Lien Pacific Group, said the Artificial Intelligence and Robot Technology Training Program was the first center in IPPG’s AI education development strategy chain in 5 years. It is expected that from now to 2025, IPPG will coordinate with universities and educational institutions to establish 10 more AI centers and deploy 1,800 AI Lab at schools and educational institutions nationwide. with the goal of AI training for more than 2.5 million students per year./.

Uran-9 combat robot is 5.12m long, 2.53m wide, 2.5m high, with a total combat weight of 12 tons, is the most typical heavy fighting robot among many ground combat robots of the Russian army . Russian Uran-9 combat robot. Source: people.com.cn. The robot is equipped with various sensor modules including laser sensor, photoelectric sensor and thermal image sensor. It is capable of automatic target recognition and tracking. It can detect and track targets to a range of 3,000 meters and maneuver in either remote or automatic mode. The maximum road maneuver speed is 35 km / h and the continuous combat time is 6 hours. The robot is equipped with 1 30mm 2A72 cannon, 4 9M120 Ataka anti-tank guided missile launchers, or 6 RPO Shmel 93mm thermo-missile launchers and 1 PKT / PKTM machine gun. At the same time, the robot can also carry 9K38 surface-to-air missiles, 9K333 mobile air defense missiles and 9M133M anti-tank missiles depending on mission requirements. It can be said that robots can create “fatal” attacks against armored targets, low-range targets and infantry forces. Looking at the development and application of weapons and smart equipment of the Russian Army, it can be seen that Russia has embarked on a path of developing military equipment integrating artificial intelligence different from the West. . Especially from the development and application of the Uran-9 combat robot, three characteristics can be seen in the intelligent military development of the Russian Army. The first is to focus on the real warfare of smart technology. Compared with the West’s search for advanced intelligent technologies, the Russian Army places more emphasis on improving its practical combat capabilities. For example, the Uran-9 combat robot is strong enough to take on an armored infantry team. Heavy armored vehicles can cope with attacks from general infantry weapons and crawler chassis can adapt to complex terrain such as hills and mountains. Especially in the actual combat test in Syria, the Uran-9 combat robot has shown a high level of attack, defense and maneuverability, which can meet the actual combat needs of the ground forces. Russian troops in such missions as fire suppression, fire suppression and infiltration of enemy front lines. In general, the results achieved by the Uran-9 combat robot are due to a design philosophy that meets the practical combat needs. Second, improve the level of intelligence based on the advantages of motorized equipment. The Russian military has always had a strongly mechanized armed force, but its intelligence still has a certain distance from the West. To quickly close this gap, the Russian Army has chosen to push the development of intelligent military equipment in the direction of modularization. The Uranus-9 combat robot’s armor, weapons and chassis are the products of the era of mechanization, while sensors, fire control systems and automated algorithms are products of the age of intelligence. . Through the combination of these two methods, Russian combat robots can quickly adapt to the battlefield in the future. Based on this method of combining traditional equipment with this advanced technology, the Russian military is developing smarter weapons and equipment. Third, to address the shortcomings of conventional forces to develop intelligent weapons and equipment. Under Western containment, it was difficult for the Russian Army to develop conventional weapons and equipment on a large scale and expand its military might. In the face of NATO’s increasingly intense military pressure, Russia needed to compensate for the shortcomings of conventional military might in a manner consistent with its economic requirements. In terms of military might, the newly formed Uran-9 combat robot unit can be said to fulfill this requirement completely. Compared to the involvement of an armored infantry army, the military costs and the risk of the robot unit’s casualties are much lower on the battlefield, which is one of the key reasons why the Russian Army plans to increase. The proportion of unmanned weapons and equipment in the military by 2025 is 30%.

Increase Armata. “With AI, Russia has created the world’s leading heavy and real robotic tanks. Armata can analyze the situation on the battlefield by himself to attack the most valuable target. After analysis, the vehicle has. can be fired automatically without the intervention of the crew or control center. What is special is that despite the integrated automatic operation, all the power and combat capabilities on the original Armata tank remain the same, “said General Manager Sergei Chemezov. According to the Russian general manager, the new Armata with standards on technical characteristics – tactics, loading system, cockpit structure and support equipment for firing control new. Armata will be equipped with a fully automated combat module that allows the turret to be used differently from traditional tanks. Meanwhile, Deputy Director of Tractor Plants Group (the unit assigned to design Armata), Mr. Mikhail Levshunov confirmed, Armata tanks will be equipped with an automated system that can automatically destroy the target. Select and it will shoot itself until the target is destroyed. He said that the remote control module research was conducted in the framework of research and manufacture of new army technical equipment models. Accordingly, the automatic system can destroy targets in the air and on land. Thanks to the heat detector, it can work all night, when it’s rainy and foggy. Especially while Russia is very confident about the combat strength of the Armata tank, the leading American magazine National Interest pointed out the backwardness of the artillery on both these tanks and the artillery of this tank. the tank front armor of modern Western tanks could not be penetrated. According to an analysis by American magazine, the main reason to say the 2А82 cannon of the Т-14 Armata is outdated is that it uses the same 125 mm size as the previous generation tank cannon. According to some sources, the artillery system was originally built to be fitted with modernized Т-90 tanks that were only intended for the Russian army. After that, with the development of the standard heavy chain chassis Armata, it was decided to equip the next generation tank. The barrel 2А82 is treated to increase elasticity and partially chromed, which greatly increases the durability of the cannon. This cannon can fire existing and future ammo types. Future ammo types could destroy not only existing tanks but also their future variants, as well as new tanks in development. Currently, Russia is testing advanced smart artillery shells for the Т-14, so the ability to destroy enemy tanks is even higher. Although the West underestimated, but the tank manufacturer Armata asserted, the barrel energy of the gun 2А82 is 1.17 times higher than that of NATO’s best cannon, the 120 mm L-55 gun. The Russian artillery is superior to all current tank guns 1,2-1,25 times in terms of technical level and penetrating ability.

Robotic dogs are trained to fight with French military trainees. According to the Daily Mail, Spot robot, priced at $ 75,000 (equivalent to 1.7 billion) was developed by Boston Dynamics (formerly a US robotics company but now owned by the umbrella corporation. Korea Hyundai as of December 2020). Close-up of robot dog worth 75,000 USD (1.7 billion VND). The robot dog is suitable for both indoor or outdoor use, can map the environment, feel and avoid obstacles, climb stairs and open doors. In particular, it can perform dangerous missions in a variety of harsh environments such as nuclear plants, offshore oil fields and construction sites or even on battlefields. Robotic dogs are trained in 3 combat missions with French military trainees. Despite being researched and developed by Boston Dynamics for many years, the robot dog was released to the market last June. One of the first customers to buy it was billionaire Elon Musk – founder of SpaceX Technology Corporation (SpaceX). The robot dog is currently one of the robots on the payroll of the French army. Robotic dogs proved to be especially useful in urban combat. “This is an expensive robot, especially for civil or police applications. New York Police (USA) are also testing it under the name Digidog. It has also been tested for anti-Covid antiseptic missions. -19 “, Clément Levilly, a robot distributor in Europe told Ouest France. The list of customers to buy this expensive robot dog includes billionaire Elon Musk. The military school Saint-Cyr has applied to robot dogs for 3 training scenarios – one is an attack mission: capture intersections; two are defensive missions during the day and at night; Finally, the combat missions in urban areas.

Country Garden restaurant (Guangdong, China) is “sold out” when the robot to cook and serve food at the table to the guests. Photo: Chinanews. The restaurant is about 2,000 square meters wide, has more than 20 types and a total of 40 robot chefs. Photo: SCMP These robots serve nearly 200 dishes. Some dishes can be processed as fast as a few seconds. Photo: Country Garden Robots can bring a feast of eyes and taste to nearly 600 diners at the same time. Photo: Country Garden The robotic restaurant serves a variety of dishes, from traditional Chinese dishes to fast foods like hamburgers and french fries. Photo: SCMP Customers can even be served the fastest robotic meal in 20 seconds after ordering. Photo: SCMP Pre-programmed robots cook the dishes according to the fixed recipe. Photo: SCMP In addition to fried dishes, the robot also quickly cooks many noodle dishes to serve diners. Photo: Country Garden After processing, the dish is delivered to the modern conveyor belt dining table. Photo: SCMP The up and down conveyor system brings food running along the ceiling area. Photo: SCMP However, robots cannot completely replace humans. For example, restaurant staff still have to bring food from robots to the table or assist with cleaning the table and cleaning it after the guests are finished eating. Photo: SCMP Video: What are the criteria for restaurants in Europe? Source: VTV24

Xinzhiyuan Report

source:Academic headlines

【Xin Zhiyuan Guide】A new generation of Xenobots-They are assembled into the body automatically through single cells, and after assembly they can be moved without the aid of muscle cells, and even demonstrated the ability to store memories.

Last year, a team of biologists and computer scientists used frog cells to create a new type of miniature self-repairing biological robots called “Xenobots”-biological machines that can move around, push heavy objects, and even make Group behavior.

Today, the team has successfully developed a new generation of Xenobots on this basis-they are automatically assembled into a body through single cells, and can be moved without the aid of muscle cells after assembly, and even demonstrated the ability to store memories. The new generation of Xenobots robots move faster, can move in different environments, and have a longer service life. At the same time, they still have the ability to teamwork and can repair themselves after being damaged.

Researchers say that this versatility makes it possible for Xenobots robots to perform tasks in more environments. Relevant research results were titled “A cellular platform for the development of synthetic living machines” and were published in Science Robotics magazine on March 31.

(Source: Science Robotics)

Shuttle at will, just because of this special “leg”

It is understood that the researchers used artificially “modified” heart cells in the previous generation of Xenobots robots. Due to the natural rhythmic contraction of these cells, Xenobots robots have the ability to move around.

In contrast, the new generation of Xenobots robots adopts a “bottom-up” approach. The extracted frog stem cells self-assemble and grow into spheroids. The cells in some spheroids will differentiate to produce cilia-like tiny hairs after a few days. Protrusions, these protrusions can move back and forth or rotate in a specific way.

The corresponding author of the paper, director of Allen Discovery at Tufts University, and professor of biology, Michael Levin, said: “We are witnessing the plasticity of cell clusters. Although these frogs have completely normal genomes, we use their cells to construct A new’body’ completely different from the default state.”

At the same time, these cells can also reuse their genetic coding hardware to achieve new functions, such as using cilia to achieve movement. Even more surprising is that they can not only take on new roles spontaneously, but also create new physical behaviors without the need for long-term evolutionary selection of these characteristics. ”

Figure | Xenobots move through cilia and show teamwork ability

In order to test the ability of a single Xenobots to move in different environments, the researchers constructed different experimental environments from a completely open field to a restricted capillary with an inner diameter of 580nm. In all cases, except for some Xenobots that were challenged in narrow passages, Xenobots were able to move in space.

Under certain circumstances, Xenobots may move in the opposite direction. The reason is that the intracellular and tissue-wide polarity is established by cell rearrangement during the self-assembly stage created by Xenobots, and cannot be changed on a short time scale. Therefore, the rapid changes observed in the study may be due to real-time control of cilia, such as changes in the beating frequency of individual polyciliated cells, rather than changes in structural polarity.

Like many unicellular and multicellular organisms that use cilia for exercise, this exploratory behavior (spontaneous changes in exercise) may be affected by random internal physiological processes and micro-changes in the environment, such as spontaneous currents in an aqueous medium.

In the experiment, when the channel width of the “maze” is reduced to 2mm, Xenobots will continuously hover around a wall of the maze. According to observations, most Xenobots tend to exhibit an elliptical movement pattern. This behavior may be due to the wall inhibiting the completion of the elliptical path, which in turn drives the final directional movement.

Figure | Xenobots can pass through a variety of terrains including mazes

In addition, the researchers placed Xenobots on one end of a 2cm capillary with an inner diameter of 580nm. Unexpectedly, 42% of the tested Xenobots were able to pass through capillaries end to end. All individuals exhibiting movement successfully passed through the entire capillaries, and none of them stayed or reversed in the middle between the two ends. To movement.

Figure | Xenobots through tiny capillaries

The researchers said that these data are sufficient to show that Xenobots can successfully traverse various environments without having to go through a special configuration to adapt to a given scene, which is an ideal feature in many software robot applications.

Stronger collection and handling functions

At the same time, the researchers also modeled Xenobots of different shapes to determine whether they exhibit different behaviors when they are alone or in a “team”.

They hope that Xenobots can do some meaningful work in real life, such as life tools for removing microplastics in the ocean or pollutants in the soil. Facts have proved that the new generation of Xenobots is faster and better than the previous generation model in tasks such as garbage collection. They can clean the petri dishes in groups to collect larger iron oxide particles.

Figure | Xenobots work together to collect piles of tiny particles

In addition, Xenobots can also cover larger flat surfaces or move in narrow capillaries. These studies also show that in the future, computer simulation can optimize the additional functions of biological robots to adapt to more complex behaviors.

Also has a memory function? !

The core feature of robotics is that it can record information and use this information to modify the actions and behaviors of the robot. In order to confirm whether Xenobots have this function, the researchers designed a Xenobots with read and write functions, and they used EosFP’s fluorescent reporter protein to record this information. This protein usually emits green light, but when exposed to 390nm wavelength light, this protein can emit red light.

Researchers injected messenger RNA encoding the EosFP protein into the cells of frog embryos. Xenobots made from these stem cells have a built-in fluorescent switch that can record blue light exposure around 390nm.

To test the memory function of Xenobots, the researchers let 10 Xenobots move on a surface where a spot is illuminated by a 390nm beam. Two hours later, they discovered that three robots were emitting red lights, while the rest remained green, effectively recording the robot’s “travel experience.”

Figure | Xenobots can record information by changing the color

In the future, this proof of the principle of molecular memory can be extended to detect and record light, radioactive contamination, chemical pollutants, drugs, and diseases. The further design of the memory function can record multiple stimuli, allowing the robot to release compounds during the stimulus, or to change its behavior based on the sensation of the stimulus.

Bongard said: “When we give robots more functions, we can use computer simulations to design more complex behaviors and the ability to perform more complex tasks. We can not only report the conditions in their environment, but also modify and Repair the conditions in its environment.”

Even self-healing

What’s more surprising is that the new generation of Xenobots is also very good at “healing” their wounds, and can heal serious full length lacerations by half the thickness within 5 minutes of injury. In the experiment, all the injured robots eventually healed their wounds, restored their original shape, and continued to work as before.

Figure | Xenobots can repair itself

The researchers said that one of the advantages of biorobots is metabolism. Unlike metal and plastic robots, the cells in biorobots can absorb and break down chemicals and work like small factories such as synthesizing and expelling proteins. The entire field of synthetic biology previously focused on reprogramming single-celled organisms to produce useful molecules, and now it can be used in these multi-celled organisms.

Like the previous generation, the upgraded Xenobot robot can survive in the embryonic energy store for up to 10 days and run tasks without additional energy. However, if they are kept in a nutritious “soup”, they can also run at full speed for several months.

Before, Levin introduced a wonderful description of biological robots in a TED talk, and he also described what we can learn from these biological robots. He not only described the great potential of micro-biological robots to perform effective tasks in the environment or potential therapeutic applications, but also pointed out the most valuable benefits of this research, using robots to understand how individual cells gather, communicate, and specialize to create greater The organism. This is a new model system that can provide a basis for regenerative medicine.

The researchers said that Xenobots and subsequent alternative versions are also expected to provide information on how ancient single-celled organisms evolved into multicellular organisms, as well as the origin of information processing, decision-making, and cognition in biological organisms.

Reference materials:

https://now.tufts.edu/news-releases/scientists-create-next-generation-living-robots

http://dx.doi.org/10.1126/scirobotics.abf1571

From Victor Gojdka, ARD stock exchange studio Manufacturing parts for customers in medical technology and the automotive industry is what Limtronik in Limburg, Hesse, specializes in. A medium-sized industrial company with a specialty: the machines in the factory could talk to each other – this is how managing director Gerd Ohl describes it: “Put simply: machine A produces, machine B checks the product I build and tells machine A: You have to Change parameters, regulate the temperature down, then I’ll have another great product in the back. ”

Many factories are “more powerful” today

Industry 4.0: the intelligent and networked factory. Thomas Gitzel, chief economist at VP-Bank, knows many companies that now operate such clever factories. “You can imagine that you walk into a factory hall where you can hardly find any employees. There is only one employee who can control the entire machine park using a kind of smartphone. That means: the software controls production and less the human with manual processes. ” Ten years ago the concept “Industry 4.0.” invented. Today, robots and the like are part of everyday life in many factories, says industry expert Volker Stoll from the Landesbank Baden-Württemberg: “In this respect, the clout of a factory ten years ago and today cannot even be compared.” Almost every fifth company in the manufacturing industry now uses robots in production. This is based on figures from the Federal Statistical Office.

Small and medium-sized businesses have a hard time

Large companies in particular work this way. But other companies are still having a hard time with it, says VP banker Gitzel: “On the other hand, we have the smaller medium-sized companies, for whom Industry 4.0. Is almost a foreign word. The small medium-sized companies simply lag a long way behind the trends after.” There are two reasons. First, converting your own factories costs a lot of money. Second: Small and medium-sized companies in rural areas often have a harder time retaining talented IT programmers. Large corporations do this better. At the Hanover Fair – the industry meeting point – the experts were now hoping for a corona effect, says Andreas Scheuerle from Deka-Bank: “There is a lot of catching up to do because companies have held back on investments in recent months. That will work out pave the way: old machines have to be replaced by new ones. ” However, not all experts are so optimistic. Because networked machines also need fast internet. And that is still the case with many medium-sized companies in rural areas – pure dream.

Xinzhiyuan Report

Source: verge

Edit: yaxin

【Xin Zhiyuan Guide】Boston Dynamics Robot Dog gets another new skill. Recently, a YouTube blogger, Michael Reeves, used his technical expertise to tune the robot dog Spot to “pee” like a real dog.

Some time ago, a woman in Florida, the United States, “walked her dog” on the beach aroused onlookers.

Because what she is holding with the “rope” in her hand is actually a Boston-powered robot dog!

Now, the robot dog Spot has also learned to “pee”, this is the real dog…

And this peeing dog was “trained” by a YouTube blogger, Michael Reeves, using his technical expertise and talented mind.

Next time you go to “walk the dog”, you must bring it, the scene is very windy! “Reminder: Be careful of getting tickets when you urinate anywhere”

Dancing, skipping rope, cleaning the house, watering flowers, going to Tesla and other factories to “move bricks”, and now Spot has learned to “pee” again.

It can be said that Spot Robot Dog got a new skill.

But the location where the outlet pipe is installed is too “evil-interesting.”

Robot Dog Spot: I am not a real person, but you are a real dog!

The robot dog Spot “urines” beer into the glass and currently has more than 7 million views on YouTube.

Let’s take a video and feel it as a whole~

The original idea of ​​blogger Michael Reeves was to make Spot the robot dog a good helper who can help people pour beer.

In the video, Michael kindly invites a friend to drink beer, but friends who see this scene should feel crazy.

“Dried this cup” really needs to overcome a great psychological barrier!

Netizens are curious, “How does it taste?”

At present, the accuracy rate of the robot dog under the supervision of Michael in pouring beer into the glass is only 35%.

Although it has not yet reached the level of “precise delivery”, this idea has already aroused enthusiastic comments from netizens.

Let the robot dog “pee” into the cup, how did Michael manage it successfully?

Training for 6 months, named “Pissbot 9000”

Michael said, “Training Spot requires some smart programming, and it took 6 months to complete.”

First, an important device is needed, a camera that can recognize where the cup is.

Since BD’s high-quality cameras cost nearly $30,000, Michael decided to use an off-the-shelf home surveillance camera, a wireless receiver and AV adapter.

Michael programmed the robot dog to find the brightest pixels recorded by the camera to identify the white interior of the red cup.

To avoid the camera’s confusion between the white ground and the white interior of the cup, Michael used the inverse square law in physics to change the point of light and project light into the surrounding environment.

The inverse square law is defined as: in any physical law, the distribution or intensity of a certain physical quantity will decrease in inverse proportion to the square of the distance from the source.

The modification that Michael had to make was not simply putting a small bucket on Spot’s back and then reprogramming.

He also designed a nozzle gimbal to aim the cup more precisely.

This device uses a servomechanism and a solenoid driven by the Raspberry-Pi “vision system” to ensure that the robot does not miss the cup.

The results of it? 35% accuracy rate.

As demonstrated in the video, “urine” sometimes spills out and messes up the floor.

By the way, this trained robot dog is no longer called Spot, and Michael named it “Pissbot 9000”.

Open source SDK, do it yourself DIY

In June 2020, Boston Dynamics opened the Robot Dog Spot to the public for public sale for $75,000.

It can walk at 3 miles per hour, climb terrain, avoid obstacles, see 360-degree images, and perform some programmed tasks.

In addition, it can also map the environment, sense and avoid obstacles, climb stairs and open doors. It can perform dangerous tasks in a variety of uninhabitable environments, such as nuclear power plants, offshore oil fields and construction sites.

In order to cooperate with the sales of this robot dog product, Boston Dynamics also open sourced Spot’s Python scripting language SDK.

You can directly control Spot with Python. Boston has open sourced Spot’s SDK, which means that after you buy Spot, you can use your own DIY functions. It is not a dream to have your own dedicated robot dog.

Just in February, a spoof organization in the United States tied a Tipman 98 paintball gun to the body of the Boston Dynamics Internet celebrity robot dog Spot, and let it wander in the gallery.

Naturally, this incident aroused heated discussions among netizens.

Boston Dynamics publicly condemned the violent, harmful, and threatening behavior for this.

Michael Reeves’s own DIY robot dog allows us to see the unknown side of Spot.

It is definitely a master who can tune a robot dog so well!

Reference materials:

https://www.theverge.com/tldr/2021/4/12/22379501/boston-dynamics-spot-pee-beer-michael-reeves

https://hothardware.com/news/boston-dynamics-spot-robot-dog-beer-delivery-pee-blast#.YHRmHyWR13Q.twitter

Nvidia subverts the CPU!Long-haired Huang Renxun enters Intel’s territory, Arm architecture CPU performance is 10 times higher

Compilation | Xiong Dabao

Editor | Li Shuiqing

Zhixi News reported on April 14 that at the beginning of this month, researchers at the University of Waterloo in Canada updated the first open source high-resolution wearable camera image database for human motion scenes. On this basis, AI and wearable cameras can be used to allow exoskeleton robots to walk autonomously.

An exoskeleton robot refers to a robot that is sheathed outside the human body, also known as a “wearable robot”. The robot has the functions of sensing, control, information, fusion, mobile computing, etc. The operator can tie it to the leg to help them enhance or restore body functions.

▲ Researchers from the University of Waterloo use the team’s exoskeleton robot to complete the walking action

1. The use of exoskeleton robots: from hand control to machine autonomy

At the beginning of the use of exoskeleton robots, it is usually necessary to rely on the manual control of the operator to switch motion modes.

Brokoslaw Laschowski is one of the main developers of this research at the University of Waterloo. He said: “Every time you want to change the way you move, you need to rely on a joystick or a smartphone app. Program to operate the exoskeleton robot, from sitting to standing, standing to walking, walking on the ground to going up and down stairs. “This is a great burden for the operator.

How to make the exoskeleton robot automatically recognize when to switch the motion mode?

▲California Institute of Technology researchers use exoskeleton robots for walking demonstrations

Scientists have attached sensors to the legs, this method can detect the bioelectric signals sent from the brain to the muscles, telling them to move.

However, this method also has some problems, such as skin conductivity will be affected when the skin sweats.

2. The accuracy rate is over 70%, and the exoskeleton robot automatically recognizes the walking environment

The Laszowski team is trying a new method: install wearable cameras for users and provide visual data to the machine so that it can operate autonomously. The AI ​​software equipped with the robot can analyze this data to identify stairs, doors and other features of the surrounding environment and calculate how to make the best response.

For exoskeleton robots to realize automatic recognition in any walking environment, a large amount of data is required.

Laszowski led the ExoNet project, the first open source database of high-resolution wearable camera images of human motion scenes. The database has more than 5.6 million images of real walking environments indoors and outdoors.

▲ExoNet project introduction interface

Laszowski pointed out that the team used the data to train deep learning algorithms. Although there are huge differences in the different surfaces and objects perceived by the wearable camera, their team’s deep convolutional neural network has been able to automatically recognize different walking environments with 73% accuracy.

At the same time, Laszowski also indicated that AI’s deep learning of images will also make it dependent on traditional two-dimensional images. When outdoor lighting and distance increase, the accuracy of AI measurement will usually decrease.

Other researchers are also using AI and wearable cameras to allow exoskeleton robots to walk autonomously.

Researchers in North Carolina installed wearable cameras on their eyes or knees, allowing volunteers to walk through various indoor and outdoor environments to capture image data that exoskeleton robots might use to observe the world around them.

Edgar Lobarton, an electrical engineering researcher at North Carolina State University, said that they focus on how AI software can reduce the uncertainty caused by factors such as motion blur or overexposure of images to ensure safe operation. “We want to ensure that before AI is integrated into hardware products, we can truly rely on it.”

3. Autopilot is inspired to improve the stability of exoskeleton robots

In the future, Laszowski and his colleagues will focus on improving the accuracy of environmental analysis software to help exoskeleton robots perform real-time operations better.

Lobarton and his team also explored how to deal with the uncertainties that motion brings to the visual system.

ExoNet researchers hope to explore how AI software transmits commands to exoskeleton robots so that the robot can analyze the terrain around the user according to the system, and perform tasks such as climbing stairs or avoiding obstacles. Laszowski said that, inspired by self-driving cars, they are seeking to develop autonomous exoskeleton robots that can complete walking tasks without human input.

At the same time, Laszowski’s team is also considering the safety of the exoskeleton robots operated by the elderly and disabled. “User safety is very important. If the classification algorithm or controller of the skeletal robot makes a wrong decision, the user must always have the ability to control it.”

Conclusion: Autonomy is the future trend of bio-assisted robots

As a branch of the robotics field, exoskeleton robots have been studied since 1960, but how to use sensors to accurately obtain the intention of the human body has always been a problem to be solved.

With the support of AI and wearable cameras, the development and landing of exoskeleton robots, bio-assisted robots, will accelerate, expanding its commercial and civilian markets.

Source: IEEE Spectrum