These Robots Are Artificially Intelligent And Have Machine Learning Capabilities


We often see and hear people talking about how advanced robotics can make factories and distribution centres more efficient. This interview looks at the effort that goes into designing these robots, through GreyOrange’s experience in designing their own IIoT products – Butler and Sorter.

Vaidya Subramaniam S, Head, Product Management, GreyOrange Speaks with Dilin Anand from EFY.

Vaidya Subramaniam S,  Head, Product Management, GreyOrance
Vaidya Subramaniam S

Q. Could you elaborate on the communications technologies used in Butler?

A. Butler possesses a variety of communications, sensing and security technologies at its core. These robots are not only move around in the warehouse based on the programmed logic but also communicate continuously with the control unit. Butler uses two wireless technologies for communication, i.e., Wi-Fi and ZigBee (open global standard). These are utilised to address the unique needs of low-cost, low-power wireless M2M networks.

Q. How do you Ensure the Butler moves about without Hindrances?

A. The sensing technology involves image processors and bar code scanners that communicate with control systems and respond automatically. For example, Butler can effectively move around at a speed of 5km/hour on a path fed into it through algorithm. It can scan barcodes and place racks at the right place through its sensing technologies. The safety is maintained in way that if anything or anyone obstructs Butler’s path, it immediately stops. Butler charges through a smart charger. It assesses the current charging level and optimises battery consumption. It utilises the central control system (CCS), to monitor the charging levels.

Q. What’s special about the way Butler moves?WHAT’S SPECIAL ABOUT THE WAY BUTLER MOVES?

A. Butler moves in a grid. It operates in a way so that the speed is optimum. Following this grid avoids any obstruction and mismanagement. Dynamic Path Planning is another feature, where Butler quickly adapts itself to the environment and can change its functioning according to the changing needs. An operator will not need to wait or sit idle, since the Butler moves quickly and effectively from Point A to Point B. So, the path of movement is effectively planned.

Q. Are there any innovative features in the optimisation and analysis engines/algorithms for Butler?

A. Butler also picks and puts the racks at the best spot so that the inventory can be accurately traced and tracked. It leaves no chance for error in inventory management. The robots are artificially intelligent and have machine learning capabilities. This means that it can understand the demand patterns. It places the racks with high demand products near the operator workstation and low demand products at the back. Therefore, warehouse space and inventory turnaround time is optimised.

Q. Moving on, what were the hardware design challenges faced while creating Sorter?

A. Sorter has moving conveyers and arms. With a lot of moving parts, there is a risk of wear and tear. The most important challenge was to ensure the prolonged life of these parts. We had to integrate a lot of functions in the Sorter system. Operators need to have a real-time view of how many products are placed over the belt, how many are sorted, and if there are any errors. These errors need to be to fixed in a stipulated period. So, the selection of drivers had to be such that it could support this process and let operators function effectively.

Q. Thought the device looks simple, I am sure it may not be from an Engineering point of view. Any peculiar challenges that you face with the Sorter?

A. At the lowest speed, the Sorter sorts 3000 packets/ hour to a maximum of 15000 packets/ hour. The product needs to be robust enough to handle such a high speed. India being a huge and diverse country, makes it a challenge to sort products basis the number of destinations/cities/pin codes, etc. If delivery must be made to 125 destinations, the sorter cannot have 125 arms to sort the packets. Therefore, we have divided the sorting into primary and secondary sorting.

Q. What are some of the deployment or Implementation challenges that you have seen in your early days, and how were they solved?

A. Developing an appropriate barcode reader was the most critical implementation challenge in our early days. Most of the warehouses in India are not air conditioned. Thus, these are exposed to dust. Our challenge was to create such a barcode reader that would not be spoiled with dust or rough cleaning. Therefore, we created a reliable barcode reader by planting a barcode protector so that dust does not stick on the scanning surface.

Another challenge came up while engineering the conveyer belt for Sorter. At a high speed, the belt often gets stretched unduly. To avoid this, we have deployed a belt made from a different material that works well on high speed and has a longer lifespan.

Q. What hardware design challenges were faced while creating these robots in India?

A. One challenge faced was to agree to the required hardware design and convert it into reality. In India, the ecosystem for product development and implementation is not as evolved as in the western world. So, the challenge was to experiment with multiple prototypes, keeping in mind that the vendors could supply according to our needs. Another challenge was to make sure that we could manufacture on a large scale. For manufacturing, getting the right kind of quality on a sustained basis is vital. To ensure this, we set up in-house quality control teams who inspect to ensure the good quality of products. Additionally, effective implementation of products at the customer site was extremely critical. The product needed to be engineered in a way that could ensure easy installation. Thus, a functional layout was made through the ‘Design for Manufacturing and Assembling’ (DFMA) process.



Please enter your comment!
Please enter your name here