Traditionally, robots have been automated objects that imitate humans or other organic lifeforms in a crude way. They can perform repetitive, difficult and boring tasks for us. But there is a lot more to come in the future.
Until recently, human-robot -interaction has been severely limited in majority of the robots. Our means of communication with robots are via programming, pressing user interface buttons and trying to stay out of the robot’s way to avoid injuries.
We want the robots to become more humane in their behavior and looks.
Because of this, one of the main goals of future robotics is to enhance the poor interface and allow us to work hand in hand with robots. While the main concern in the design is fluent controllability, we want the robots to become more humane in their behavior and looks.This enables us to work comfortably with the robots in both industry and service tasks. Ultimately, we will integrate robotics permanently to our society and way of living.
We are expecting to see self-driving cars, drone deliveries and companion robots. It’s also likely that most blue-collar jobs such as public transport will be done by robots as soon as the technology has reached that level. Needless to say, the largest and oldest industry in the world, food production, will be facing a revolution as it will be automatized. That’s what we are going to talk about next.
Automatizing the fight for survival
The global population is expected to reach 9 billion by 2050. According to IEEE Robotics and Automation Society agricultural production must double to meet the future demands for food. Robotics is the most prominent way to solve this problem.
As discussed above, robots do repetitive, difficult, dangerous and boring tasks for us. Agriculture is full of those. In fact, agriculture ranks among the most hazardous industries. Farmers get exposed to the chemicals used in fertilizers and pesticides. Even worse, tractor overturns are the most common cause of fatalities in agriculture.
The modern human-operated tractors have another problem: they are huge. Even though efficiency can be increased with larger units the trade-off is the tractor’s effect on the soil. Massive tractors compact the soil. A compacted soil has a reduced rate of both water infiltration and drainage. Future robots could avoid this effect with several lighter units collaborating in a swarm. The swarm is a fault-tolerant system as failure of a single robot does not necessarily cause failing the task. The other robots may even provide service for the broken one.
Why have the agricultural companies not yet replaced majority of their workers with robots?
We have already robots that can do spraying, weeding, and seeding. Why have the agricultural companies not yet replaced majority of their workers with these 24 hours a day working machines that never get hungry, tired, sick or distracted? The problem is that the current robots are specialized to certain tasks and environments. Thus, they do not replace enough human labor.
For fully autonomous operation, the robot should be able to refill its energy source and possible resource supplies such as seed containers or pesticide tanks. Also the robot should handle the numerous situations where it has confronted some error or malfunction and needs service without human intervention. Considering this we are far from a system where the whole chain from buying and collecting supplies to the actual problem solving and transporting the harvest to warehouses is done without humans. Regardless of the challenges, that is the goal of future robotics and it will be reached sooner or later.
There will be consumer-grade farming robots in the future as well. FarmBot is an open-source farming robot for autonomous vegetable production in a backyard. It makes local food production easier and guarantees environmentally friendlier and healthier food and enhances the market for those. As such technology lowers the threshold to begin a business there could be a shift from the predominant huge companies to smaller food producers. Future agricultural robotics might thus begin a shift to more responsible food industry.
The story of Unicorn
I have got my hands dirty with agricultural robotics. And when I say dirty, I mean dirt-y. Our team of eight students and two instructors built a robot, UniCorn, which does agricultural tasks such as navigation and weed spraying autonomously. The motivation was a university course that aims to learning project work and management in a team through a real-world challenge. Our showdown was at the 14th annual Field Robot Event which was held this year in Harper Adams University, UK.
One of the most time consuming tasks in the competition was advanced navigation, which we ended up winning. In the task the robots were required to navigate a pre-defined path in a maize field. The path included different turns and driving in the maize rows without damaging the plants. The task was not trivial: many teams ended up interfering in the robot’s performance because of a technical failure.
Positioning and navigation are the most fundamental parts of robot operation. Without such systems the robot does not know where it is. In a way, all other tasks are thus dependent on navigation: the robot can’t spray a weed successfully if it doesn’t know where the weed is relative to its sprayer.
Our robot had several sensors for getting information of the world: a SICK lidar, four ultrasonic sensors, a gyroscope, encoders in the axle modules and machine vision system. With these sensors and the related software algorithms the robot could get the distance of all objects around it and avoid collision with them. This is what navigation is about.
The algorithms were based on a state machine that has different actions for each state: what to do when the robot is trying to avoid collision or when it is turning. The turn had even its own state machine as it was composed of many parts. All actions were finally broken down to sending messages like “turn the wheels” or “drive forward” to the axle modules.
Now, we have come a long way from robotics to the actual problem solving in a specific task in navigation. Fundamentally, all problem solving comes to gathering data and processing the data in order to act accordingly. With clever algorithms you can create robust solutions without expensive sensors. This is what motivates me to work with robotics.
End of the article, I’m off to hang out with my robot
Autonomous technology has the power to make our life easier. It will also change the way we live. Much of the future depends on the evolution of robotics as it will be the main technology used.
I enjoy doing robotics because it blends mechanical engineering, electronics, programming and problem solving in an intriguing fashion. Why should you do robotics? Well, go on and decide for yourself whether it’s the tech, fun or changing the world. Either way, you’re going to have a good time.
The author is a robotics master student in Aalto University.
This piece starts a new series on our blog: in-depth articles and guest blogs of all things tech. Is there something you’d like to know more about? Comment below your ideas and we’ll make sure it’s featured on Junction’s blog this coming fall!