into hanchine-月博登录中心入口

in the united states, the unemployment rate has fallen to its lowest level in nearly 50 years, with 38% of manufacturing labor moving parts between boxes and machines, but there are still 500,000 jobs vacant. people urgently need a random bin picking system (random bin picking) to deal with the problem of labor shortage, and many suppliers in the automation industry are striving to meet this demand. in recent years, every automation equipment trade show will add several new companies, claiming that they have finally solved the problem of disorderly crawling[1]。

random bin picking parts from a box (random bin picking) and accurately place them in the machine. this is a simple task for humans, but a difficult challenge for robots. the robot must go deep into the corner of the box and be able to grab parts from countless directions while avoiding collisions with the box, other parts or the work unit itself.

a disorderly grasping system must include 3d vision imaging and point cloud analysis, hand-eye calibration, collision detection, grasping planning, motion planning and other technologies. our company has shared this aspect. for details, please refer to the article "han zhen sharing | robot how to quickly and accurately grasp disordered and scattered workpieces? ——3d visual target recognition and positioning".

the basic composition and implementation process of a 3d disordered capture system are shown in the following figure.

a lot of integration and programming work is required to implement such a disordered grasping system, so most of the disordered grasping systems are deployed in large and complex manufacturers' factories (such as automobile original equipment manufacturers). however, the labor force of small and medium-sized enterprises accounts for 69% of the global industrial labor force. they have the most labor shortage and need a disorderly grasping system more than large manufacturers. however, they face the problem of insufficient funds and professional skills.[1]。

so what kind of disorderly crawling solution should be provided?

first of all, a truly universal out-of-order crawling solution needs to be used by non-professionals and can be configured within a few hours;

secondly, this system needs to provide stable and reliable 3d vision recognition positioning, collision detection and robot path planning algorithms, and it can work with little or no tuning;

finally, it is low-cost, so that more companies, especially small and medium-sized enterprises can choose to use it.

as a user, the ideal process for purchasing and deploying a set of disorderly crawling system should be as shown in the figure below.

to this end, our company has launched binpicker 3d intelligent capture software, which integrates 3d point cloud recognition algorithm, hand-eye calibration algorithm, collision detection and motion planning algorithm. at the same time, our company’s vision controller can support multiple adaptations at home and abroad. branded 3d cameras and robots provide one-stop solutions (the price is one-third of the same type of foreign products).

the company's r&d personnel have done a lot of exploration on the ease of use and compatibility of the product, and finally we launched a five-step solution to complete the disorderly grab configuration. the steps are as follows.

for different application scenarios, our company is also equipped with a full range of 3d cameras, including trued2305m and trued2305m for detection, and trued2305l and trued2303xl for disordered capture.

some people may ask, since you provide such a cost-effective and easy-to-use solution, can you announce that you have solved the random bin picking problem? the answer is no, because there are too many complicated working conditions that we can't solve, and we still can't really replace labor in the foreseeable period of time in the future.

for example, "hanchine sharing | how can robots quickly and accurately grasp disordered and scattered workpieces? ——3d visual target recognition and positioning mentioned in ", some people call rbp a holy grail in the field of robotics applications. we are just one of many challenges to this problem. although many rbp systems have been put into commercial use in the world, domestic start-ups have also made good progress. many applications have been able to operate stably in poorly-conditioned factories for a long time, but they are far from "mature".

as an explorer of 3d vision applications, hanchine is committed to creating more, smarter, and more economical products to help chinese manufacturing companies to upgrade their intelligence.

references：

[1] eric truebenbach. is fully automated bin picking finally here?[db/ol].https://www.therobotreport.com/fully-automated-bin-picking-finally-here/  2019