Estevan Silveira @ Envisioning
The history of robotics is intertwined with the history of automation. Aristotle (384-322 BCE), in his Physics, dreamed of tools that could, when ordered by humans, work on their own, as if "looms were to weave by themselves." A longing that definitely became reality centuries later with the rise of industrial automation processes and, in more recent times, thanks to robotics, fully operational automatons. Robotics is an interdisciplinary domain detached from computer science and mechanical engineering centered on the building of machines ―hereinafter called robots― that can replicate or even replace human actions.
The world of scientific fantasies are abuzz with robots overtaking humans or, alas!, their jobs. Before robotic takeover strikes, let us look at this in a reflective way. 'Taylor' is the problem, not 'Ford'. That is, we should be more concerned with robots occupying the placements of our bosses, not the shop-floor: goods-to-person and collaborative picking robots are welcome, but there is something fishy about thorough managerial automation, the hyper-rationalization of command and control. Think about it for a second: how can something as subjective as leadership be automated?
Eliza, the mother of all chatbots, or conversational agents, was created by Joseph Weizenbaum at MIT in 1966. Her job was to establish convincing dialogues with humans, to the point that they could not determine whether she had robotic origins, something like the Voight-Kampff test from the movie Blade Runner (1982), designed to distinguish androids from humans. When that happens,
regardless of the outcome, you might have to realize that robots are hardware plus software. Bots are algorithmic robots coded to automate conversation.
We can say that, up to a point, chatbots are a combination of social robotics, a term usually associated with Robot Caregivers, and service robotics, a term usually associated with housekeeper robots. More precisely, chabots can be one of the touchpoints that are bundled as part of a customer journey, in accordance with a service design or omnichannel strategy. Maybe you would like to merge usual chat programs with GPT-3, a new AI language model that employs big deep neural networks. But when it comes to semantics, robots begin to create authoritative statements that are completely false. Chatbots, even if AI-powered, have hard times reading between the lines.
Robotics Education Ecosystems
Some years from now, wherever your eyes land, you will always find one of them. Welcome to Robotland. Fully autonomous, or 'self-driving' electric vehicles, run on roads able to charge batteries on the fly through Inductive Transpor Charging. Boston Dynamics' fleet, formerly frightening, now gently co-work with humans in warehouses to retrieve goods. Kuka's industrial robots have left behind doing only welding-type tasks and began to perfect themselves to 4D printing active origami. And so forth.
Today, things are more subtle. As imagination retreats, other realities appear to be so much more urgent, bringing our feet back on the ground, as the use of robots in education. To spice the robotic pull force exercised on children, learning model policies should add an 'a' to the STEM acronym. So, Science, Technology, Engineering and Math becomes STEAM, or Science, Technology, Engineering, Arts and Math. It is worth noting that some visionaries are building a robotics education ecosystem across Africa, combining video game characters, robotics and AR. Presumably, the Global South has a plan to leapfrog the North.
What Lies Ahead
The so called end-effectors have been seen as ripe for innovation. Those are devices that attach to a robot's wrist, allowing it to pick up and place materials of various shapes, sizes, and textures. Eventually, they may forgo solid things on behalf of more delicate or biological materials. There's so much development going on in soft tissue surgical robots, flexible manufacturing in the food industry, and, laying grippers aside for a while, deep-sea exploration. Speaking of the sea, think of robots made of elastomeric polymers, as Robert F. Shepherd's multigait soft robot, inspired by marine animals that do not have hard internal skeletons. Look now to the future, in which nature-restoring biobots, in the form of aquatic plants, clean unwanted algae off dead coral, produce medicine seed pods for fishes and absorb excess carbon dioxide, helping relieve ocean acidification. In the matter of climate change remediation, soft is better than hard.
The next step for centralized multi-robot systems is the development of robot swarms, that is, decentralized units souped-up by sensing autonomy and self-organization. At this stage, we have simultaneous localization and mapping (SLAM) robots, whose main job is exploration and production of accurate maps. In due course, we shall see robotic devices with a certain degree of epistemic autonomy capable of modifying their own internal structure in open-ended ways. Therefore, multi-robot SLAM will not only share in a coordinated way raw and processed data, but will be fitted with evaluative parts that directs the modification of the mechanisms that mediate sensorimotor coordination of each individual.
This energy transfer mechanism can be applied to roads and highways and is able to wirelessly charge batteries of onboard vehicles by creating an alternating electromagnetic field with an in-vehicle induction coil. By installing primary coil modules within the road surface, a magnetic field is created, which generates an electric current in a secondary coil placed under the vehicle, powering the vehicle's batteries. Some initiatives are studying magnetizable concrete materials to achieve inductive transport charging. This solution functions with parked vehicles using fixed pods or moving vehicles in a ‘dynamic charging’ mechanism through electrified roads. Additionally, wireless charging could become bidirectional: not only from the road to the vehicle but vice versa, by harnessing the energy generated from braking.
A method through which a computer digitizes an image, processes the data and takes some type of automated action. Machine Vision allows systems to understand and interpret the environment using live or recorded images, tag their content, and enable programs to perform automated tasks that previously required human supervision. By using one or more video cameras with analog-to-digital conversion and signal processing, a computer or robot controller receives the image data. This also allows systems to perceive their surroundings beyond regular electromagnetic wavelengths and might include infrared, ultraviolet, or X-ray frequencies to enhance image processing precision.
DASH: Digital Access to Scholarship at Harvard
DASH: Digital Access to Scholarship at Harvard