Mobile robots and autonomous vehicles in intralogistics delivery and implementation

Mobile robots and autonomous vehicles in intralogistics delivery and implementation

24/01/2021

Automation in the supply chain, logistics and warehousing and implementation is a rapidly growing market. Especially when it comes to using mobile robots, drones and autonomous vehicles to automate motion-based tasks.

AUTOMATION OF INTRALOGISTICS

This field covers a diverse range of robots, drones and autonomous vehicles that assist goods in their journey from storage to destination. For example, the trade in electronic components is changing the way warehouses are built and operated. What do we mean by this? That warehouses need to become more efficient in handling the immediate fulfillment of orders across multiple items. The use of automation is therefore an important part of the answer to this requirement.

Automated carts / AGVs have been in use for quite some time. They are infrastructure dependent which means they follow a fixed infrastructure such as a cable placed under the floor which creates a magnetic field or a magnetic tape as it travels from A to B. Therefore they are reliable to handle any type of load.

However, their installation is time consuming. However, in their classic forms, they are neither quick nor effective users of space. They are not suitable for cooperation in the human-robot workflow.

As technology, they are on shaky ground unless they adapt. This is because technology is evolving towards a navigation that is more autonomous and infrastructure independent.

A BRIGHT PLACE FOR AUTOMATED ROBOTS

The bright spot in the whole of this situation for automated mobile robots is goods-to-person automation in logistics centers. In warehouses, special zones are created exclusively for robots, in which these robot fleets move the racks at high speed to the picking station. The performance increase is clear and also proven. The technology certainly saves space and ensures consistency in warehouses.

Many product design innovations have helped this market grow. The equipment requires good acceleration and deceleration to operate at high speeds flawlessly. The racks require special adjustments to keep the load stable during transport. The suspension system – which raises and lowers the racks – requires special design and engineering. The robots are equipped with multiple motors which give them many degrees of freedom of movement. However, the navigation technology itself is not complicated.

The key value-adding technologies, however, are on the software side. This covers the entire stack, from the custom firmware found in the engine drivers to the fleet and job management levels. It is a rapidly expanding market space, and it all shot up into the air when Amazon acquired Kiva Systems in 2012, thus leaving a market gap.

THE FUTURE TOWARDS AUTONOMOUS ROBOTS

The outlook for classic AGVs does not look optimistic. The main reason is that navigation technology is moving from automated to autonomous. The primary benefit is that navigation becomes infrastructure-independent, which allows you to easily modify your workflow. Autonomous mobility also enables different modes of cooperation between robots and humans, thus extending the usefulness of such autonomous mobile robots and vehicles.

Current technologies enable better algorithms based on different sensors, including stereoscopic cameras and 2D laser scanners – the devices are sufficiently developed to support safe autonomous navigation in many structured indoor environments with a high degree of control and predictability. These robots are easy to install and train.

However, there are still many technological options and choices have long-term strategic ramifications. A common process is the use of 2D laser scanners to develop a map of the objects during the training phase, e.g. the movement of the robot over the objects. The fixed reference objects will be selected during the setting stage. This system is quite simple. However, it does not do well in dynamic and changing environments. Another approach is to use camera vision and deep learning to identify and classify objects as well. This is more computationally complex, but will enable a more flexible system that can have more intelligent decision-making in a complex and changing environment. It extends the utility of such vehicles to more internal scenarios by enabling the mobile robot to respond to a changing environment.

The business models of the suppliers of these machines also vary and evolve. Some offer their technology as RaaS (Robot as a Service). The point is that users don’t need robotics expertise, don’t require upfront capital, and don’t have to worry about the risks of obsolescence and technological change. The model also fits within the users’ operating budget, further facilitating adoption.

➡ Why is the RaaS model one of the fastest growing in the industry? Because the model allows the company to react quickly to the current market needs, giving unprecedented flexibility.

On the other hand, many companies follow the traditional hardware sales model. Here, providers are forced to build a platform for their business model to offer maintenance and updates, especially software updates in the cloud.

The market of autonomous mobile robots (AMR) will surely grow. The forecast in many reports shows that in the years 2020-2030 over 200,000 units may be sold. robots: this number includes robots that can assemble regular or irregularly shaped pieces.

THE AUTONOMOUS FUTURE OF FORKLIFTS AND TUGS

Forklifts and tugs are an indispensable tool in warehousing and logistics. They perform many functions. Today, almost all forklifts are manually operated. However, the development of autonomous mobility technology is slowly changing this trend. The journey has long started, and many manufacturers have already developed, demonstrated and implemented autonomous forklifts or tugs in the logistics departments of many companies.

The choice of navigation technology varies. Some people use RGB cameras and RGB image processing technology to navigate. In the past, this field was extremely difficult. However, advances in deep learning have completely transformed the field in the last 6-7 years.

Current technology enables excellent location and recognition of 2D objects, even surpassing human-level capabilities. Error rates in the recognition of 3D objects are still high, and most importantly, they will evolve rapidly, especially when using sensor fusion, e.g. involving data from laser scanners. Camera-based technology will provide a more complete long-term plan to improve navigation. Many of them still use a 2D laser scanner as it is easier to implement and is often good enough in a known, controlled and slowly changing environment.

The cost of these forklifts is obviously higher, but many vendors have a declared ROI of 12-18 months.