Historically, the larger robotics sector has been beset by several issues including the following:
- Low Return on Investment (ROI) – Due to low costs of labor, the need for extensive integration of robotics in fields such as manufacturing and agriculture has been slow.
- Issues concerning dexterity – There is a limitation in mechanical capabilities as well as robot gripping.
- Problems with vision – Most industrial firms heavily rely on cages to protect humans from robots that can’t move around them. Therefore, vision technology that allows robots to identify and navigate around objects, as well as people, have been slow to improve.
However, ever since the introduction of collaborative robots about ten years ago, things have started to change as some of these challenges are being tackled. Below is how companies and businesses are using collaborative robots to address these technological issues.
Collaborative robots leveraging new technology
As industrial robots swarmed into factories all over the world, visions systems that allow the machines to identify and navigate around objects and people were mainly an afterthought. Since a plethora of robots carry out repetitive tasks in environments that are highly structured, safety measures have typically consisted of a cage which is used to prevent human workers from getting dangerously close to the machine. However, this tide is currently turning.
During the last couple of years, vision hardware has not only become cheaper and more effective but also more widespread. For this reason, most startups today prefer to work with vision-enabled robots that are laser-laden and can operate alongside human workers without having to be confined to a cage.
Having been invented in sometime around 1996, cobots now come in an array of different shapes and sizes. They are also designed with human workplaces in mind. Even though they are relatively autonomous and easy to program, collaborative robots still lack the strength of low-tech industrial robots.
Collaborative robots are disrupting the robotics industry because they can easily complete predictable and repetitive tasks. Where less structured tasks such as picking an item from a random assortment are concerned, there exist more unique scenarios where special handling needs to be calculated by an algorithm. Therefore, if AGVs are to become truly collaborative, they must autonomously see obstacles and act accordingly.
Where we are now
Even though teaching a robot to map and manipulate objects has proven to be a cumbersome job, there are research developments that have demonstrated success with one-shot learning where collaborative robots can recognize new objects without much data training. In the future, robots may only need to watch a human once to learn a task, or they could even be programmed using VR gesture control.
While this tech still requires a few more years of research and design, the systems being used today are already good enough for collaborative robots to work alongside human colleagues. At present, cobot arms together with automated guided vehicles (AGVs) have become standard in both warehouse and manufacturing environments.
Collaborative robots seem to have found a sweet spot especially within smaller factory operations where human labor is still required such as completing more cognitive tasks including picking warehouse orders’, manufacturing medical devices or tending to 3D printers among many others. However, there is still plenty of room for the improvement of robotic vision.
Overcoming mechanical challenges
Dexterity still remains a major challenge in the field or robotics. However, there is hope because hardware improvements are making collaborative robots more effective. Because of better and cheaper hardware, startups are now focusing on tailoring software for specialized tasks as well as computer vision. Universal Robots which is the largest manufacturer of cobots at the moment features robotic arms that serve as the pick and shovel for a new wave of startups. Some arms pay themselves within an average of 195 days. For more information on collaborative robots go to universal-robots.com.