BAXTER, the collaborative bot

        The mid-level and small industries all over the word have a lot of mundane tasks which do not require speed and precision. These tasks include jobs like loading, unloading, sorting, etc. The industrial robots have a good deal of speed and accuracy but for a specific task and hence not much desirable by these industries. Also the high cost of industrial robots is one of the factors which cannot be ignored. So, these tasks are to be automated. This gave birth to the idea of “COBOTS” or collaborative bots. A collaborative bot is basically a robot which aids or guides or learns to do a repetitive task alongside human workers. These tasks which are performed repetitively by humans tend to make them – as they say “robots”. The two main obstacles for use of robots in industries are ease of use and cost. If these barriers are crossed, then the entire manufacturing industries will be revolutionized. Manufacturers have been trying for a long time but had little success in having a wide range of cost-effective operations for jobs in production which are high mix and low volume. If an industry integrates a cobot to their workforce, they will allow the workers to improve the quality of their product and hence make them gain a competitive advantage over others.
          
        A cobot named BAXTER is the main theme of this article. Baxter was designed and built by Rethink Robotics founder Rodney Brooks and his group of mechanical and signal engineers. Five years of secretive development and US $62 million in funding went into effort before Baxter became a reality. The funding providers were Bezos Expeditions and Charles River Ventures. Brooks’ vision is that robots would be so inexpensive that every factory would be able to afford one.
Baxter is designed to do repetitive stuff like picking something up, putting it down somewhere else while at the same time adapting to changes in the environment. The workers are not at risk when Baxter is at work alongside them and hence is inherently safe. Baxter is neither particularly fast nor precise but speed and precision is not required to that great an extent for a lot of jobs. It has a limited speed and lower weight (79 kgs). It has an artificial intelligence or AI computer which allows it to learn tasks from a person which need not be a professional in robotics. Any person can program a new task simply by moving Baxter’s arms and following the prompts on its LCD menu which also acts as its face. As compared to industrial robots having the same features which will set an industry aback for hundreds of dollars, the Baxter is available at just $25000. But, how is this achieved? How is the cost so low as compared to the other robots? The answer lies underneath. Thousands of Baxter’s parts were ingeniously engineered and materials procured from local sources to keep the cost of its building under check. The other way the cost was cut down was through its software. The software allowed Baxter to autonomously compensate for its own mechanical irregularities as well changes in the environment which prevented use of costly components.
As per Brooks, there are about 300,000 small and mid-level manufacturers who can afford a Baxter or two. Also he pointed out during the release of Baxter that outsourcing manufacturing to China was not sustainable as the cost of Chinese labor
goes up for the industries, the appeal of doing a product there would soon go away. He said that a simple robot could do a lot of tasks in the factory which includes basic material handling, packing and unpacking boxes, polishing and grinding. Also, the advances in processors and sensors that were making PCs and smartphones better and better could easily be integrated to make the robots better.

          All in all the Baxter has one key feature: compliance i.e., it is flexible and can control the forces it applies to things. Robots like Baxter but not exactly similar have been made by robotics industries like Redwood Robotics, Kawada Industries, Adept Technology, Barrett Technology, etc. Rethink Robotics will open the platform to third party software developers which will make Baxter do a lot of tasks which are not envisioned yet. On doing so, the robot will become more efficient and could very well replace costly human workers. A similar version of Baxter could be seen at McDonalds flipping burgers, pouring coffees in big coffee outlets, folding T-shirts at a big cloth store like Levi’s and the list is endless.

Comments