Grigory Bolotin, Chairman of the Board, "Engineering Cluster" Association
Mikhail Bolotin, Expert in Developing Innovative Models of Industrial Equipment
Andrey Gitsenko, Expert in Industrial IT Solutions Development
Download in PDF format the article "Are Russian Technologies Ready for Automation of Construction Equipment?"
The concept of unmanned construction equipment: tasks and features.
The concept of Unmanned construction equipment has been on the agenda of technology forums for the past 15 years. Public attention is typically focused on road vehicles. All major global technology companies, including smartphone manufacturers, have established divisions to create their own self-driving cars. The main challenge for autonomous vehicles is safety on public roads, and the discussion about legal aspects is far from over. At the same time, there is an entire market where millions of units of equipment operate 24/7 in a closed environment with minimal human presence – construction equipment. Can autonomous vehicle technologies be transferred to construction machinery? When are we going to see bulldozers, excavators, and dump trucks working autonomously, without human intervention, to dig a foundation pit for the next skyscraper or trenches for a multi-kilometer pipeline?
It is worth noting that autonomy and electrification are two competing trends that have long competing for a key role in the technological agenda. However, these technologies often go hand in hand, as electric-powered machinery is easier to automate. Projects developing autonomous construction machinery became widespread in early 2000s. Yet, the technology progressed from mock-ups to prototypes by the early 2010s, when serial components for automation and machine navigation became more accessible, and engineers with experience in autonomous vehicles joined the game. Electrification of construction equipment is primarily developing through the hybridization of large machinery, where an internal combustion engine (ICE) on board acts as a generator. Such machines are driven by electric traction motors.
Before delving into the technical details of construction equipment automation, it is essential to highlight the key differences between the tasks of autonomy in off-road and specialized machinery versus those in the automotive sector:
An excavator must execute the correct movements of its boom and bucket to dig a trench as specified in the project documentation. A bulldozer must not just move itself from one point to another but also set the correct blade tilt angles and height to move materials according to the project task. A wheel loader must position its bucket correctly based on several parameters to capture the maximum allowable amount of material without overloading. A dump truck must position itself for loading in precise coordination with the actions of an excavator or loader, and its unloading point is constantly shifting.
From the above, several basic requirements for autonomous construction equipment technology can be formulated.
The technology for transitioning to unmanned construction equipment must be end-to-end, integrating related technologies that will allow:
It is worth noting that autonomy and electrification are two competing trends that have long competing for a key role in the technological agenda. However, these technologies often go hand in hand, as electric-powered machinery is easier to automate. Projects developing autonomous construction machinery became widespread in early 2000s. Yet, the technology progressed from mock-ups to prototypes by the early 2010s, when serial components for automation and machine navigation became more accessible, and engineers with experience in autonomous vehicles joined the game. Electrification of construction equipment is primarily developing through the hybridization of large machinery, where an internal combustion engine (ICE) on board acts as a generator. Such machines are driven by electric traction motors.
Before delving into the technical details of construction equipment automation, it is essential to highlight the key differences between the tasks of autonomy in off-road and specialized machinery versus those in the automotive sector:
- Construction equipment operates in conditions of a constantly changing landscape. The terrain is continuously being modified, which prevents the use of proven automotive solutions, such as road graphs. Even quarry roads, which may seem somewhat similar to certain types of automotive roads at first glance, can change within a few days during intensive operations.
- Construction equipment operates within confined sites with minimal human presence, which significantly simplifies, but does not completely eliminate, safety issues.
- Construction equipment performs significantly more complex tasks than road vehicles. The task of a road vehicle is to move from point A to point B on a flat road with markings and signs, while obeying traffic rules and avoiding collisions with other road users. In performing their tasks, construction machines have far fewer constraints but operate under more complex conditions, as mentioned above.
- Another crucial difference is that each type of machinery performs a specific kind of work, which fundamentally differs from others in terms of kinematics, use of working tools, and task metrics. Instead of simply moving from point A to point B, albeit with many constraints, construction equipment must not only move around a site with constantly changing terrain but also alter that very terrain using its working tools. Moreover, the terrain can change significantly throughout the day, making it impossible to use common terrain maps as in the automotive sector.
An excavator must execute the correct movements of its boom and bucket to dig a trench as specified in the project documentation. A bulldozer must not just move itself from one point to another but also set the correct blade tilt angles and height to move materials according to the project task. A wheel loader must position its bucket correctly based on several parameters to capture the maximum allowable amount of material without overloading. A dump truck must position itself for loading in precise coordination with the actions of an excavator or loader, and its unloading point is constantly shifting.
From the above, several basic requirements for autonomous construction equipment technology can be formulated.
The technology for transitioning to unmanned construction equipment must be end-to-end, integrating related technologies that will allow:
- (a) continuously monitoring of changing terrain surface and working conditions,
- (b) formulating and executing control signals for different types of equipment based on a high-level human-defined task,
- (c) having a centralized task assignment unit to ensure interaction between different types of equipment and high efficiency of the work execution chain.
What has been done, domestically and worldwide?
All projects in the field of autonomous or automated construction equipment should be considered from the perspective of creating a high-level technology that solves the three tasks outlined above.
Task 1.
First priority Task 1 - continuously monitor the changing landscape and working conditions – is essentially solved through the application of two technologies: 1.a) the use of UAVs (Unmanned Aerial Vehicles) for construction site monitoring and 1.b) the use of computer vision systems and laser radars directly on the equipment itself.
The technology of using UAVs for construction site monitoring, in turn, consists of two related tasks:
In the context of the overall task of equipment automation, solving Task 1 will allow for frequent updates (if necessary, up to once per hour) of high-precision digital elevation model. Obtaining such information will not only allow for optimizing routes for equipment movement across the site but will also create conditions for generating work assignments to modify the terrain surface for the direct operation of each type of machine.
The technology of using UAVs for construction site monitoring, in turn, consists of two related tasks:
- Territory scanning using UAVs. Various types of UAVs (e.g., DJI, Russian manufacturer Geoscan) as well as automated UAV flight systems, so-called "droneports" (also DJI, Russian manufacturer Rusdroneport), are widely available both globally and in Russia.
- Processing and analyzing obtained raw data to create and work with a digital twin of the construction site. Solutions for these tasks are also widely available. Besides Western companies like Pix4D and Propeller, Russian developments, such as the Skyeer IT platform, are also available and widely used.
In the context of the overall task of equipment automation, solving Task 1 will allow for frequent updates (if necessary, up to once per hour) of high-precision digital elevation model. Obtaining such information will not only allow for optimizing routes for equipment movement across the site but will also create conditions for generating work assignments to modify the terrain surface for the direct operation of each type of machine.
Task 2.
The key step from remotely controlled equipment to robotic equipment lies in the functionality of generating control signals for moving the equipment and operating its working tools. Moreover, these control signals can be different for each type of equipment.
The basis for implementing this functionality is machinery that has actuators for its working tools and can be controlled remotely by a human operator. Such machines have already been developed, tested, and are available both abroad (e.g., various equipment from Caterpillar, Doosan) and in Russia (e.g., DST-Ural bulldozers). This equipment must also be equipped with an integrated telematics system that allows monitoring all operational parameters, not only of the machine's aggregates but also of its working tools, to obtain feedback and fine-tune control programs in real-time.
The functionality of Task 2 must consider both the terrain data obtained from solving Task 1 and the parameters of the equipment's working tools, as well as the overall process goal. Based on this information, an individual work task for the equipment should be created for each specific situation. Elements of the logic for such algorithms are embedded in so-called decision support systems – digital operator assistants, such as auto-steering systems and machine precision positioning systems.
In global practice, full-fledged solutions for equipment automation have been implemented only for specific types of highly specialized machinery. For example, the company Built worked for several years on a project to automate excavators as the most widespread type of construction equipment. However, in 2022, the company decided to focus on selling automation solutions for piling machines based on excavators. This example confirms the complexity of implementing this functionality, even as experimental prototypes.
Honda has been testing autonomous cargo carriers for a long time. Similar developments are being tested by Volvo and Doosan. In Russia, the current leader in this area is truck manufacturer KAMAZ, which is testing an autonomous unmanned dump truck this year. Another is the company Cifra, which has created several units of autonomous quarry dump trucks. However, these developments have not yet seen serial application.
The basis for implementing this functionality is machinery that has actuators for its working tools and can be controlled remotely by a human operator. Such machines have already been developed, tested, and are available both abroad (e.g., various equipment from Caterpillar, Doosan) and in Russia (e.g., DST-Ural bulldozers). This equipment must also be equipped with an integrated telematics system that allows monitoring all operational parameters, not only of the machine's aggregates but also of its working tools, to obtain feedback and fine-tune control programs in real-time.
The functionality of Task 2 must consider both the terrain data obtained from solving Task 1 and the parameters of the equipment's working tools, as well as the overall process goal. Based on this information, an individual work task for the equipment should be created for each specific situation. Elements of the logic for such algorithms are embedded in so-called decision support systems – digital operator assistants, such as auto-steering systems and machine precision positioning systems.
In global practice, full-fledged solutions for equipment automation have been implemented only for specific types of highly specialized machinery. For example, the company Built worked for several years on a project to automate excavators as the most widespread type of construction equipment. However, in 2022, the company decided to focus on selling automation solutions for piling machines based on excavators. This example confirms the complexity of implementing this functionality, even as experimental prototypes.
Honda has been testing autonomous cargo carriers for a long time. Similar developments are being tested by Volvo and Doosan. In Russia, the current leader in this area is truck manufacturer KAMAZ, which is testing an autonomous unmanned dump truck this year. Another is the company Cifra, which has created several units of autonomous quarry dump trucks. However, these developments have not yet seen serial application.
Task 3.
As with any construction project, a fleet of robotic equipment must have centralized management that ensures timely task assignment for each unit, machine-to-machine interaction algorithms, and synchronization of these tasks to build an efficient end-to-end production process. A key feature of this functionality is that the algorithm must have feedback with the algorithms performing Task 2 and constantly update information about the construction site status within Task 1.
Implementing such algorithms for managing a group of machines involves elements of artificial intelligence, as the project implementation will accumulate a large amount of diverse data: construction site status, condition and operating modes of the equipment and its working tools and coordination of multiple machines. This data will need to be consolidated into datasets and used, among other things, as tools for so-called decision support systems.
Implementing such algorithms for managing a group of machines involves elements of artificial intelligence, as the project implementation will accumulate a large amount of diverse data: construction site status, condition and operating modes of the equipment and its working tools and coordination of multiple machines. This data will need to be consolidated into datasets and used, among other things, as tools for so-called decision support systems.
Project Horizon
The Horizon Project, which we plan to launch in 2023, aims to implement and refine the functionality for solving Task 2 (generating control signals for various types of equipment), with a subsequent transition to creating the universal intelligent Horizon Platform. This platform will allow combining different types of equipment into a single production chain and will represent a unified center for monitoring and managing a fleet of construction equipment, thus implementing a solution for Task 3.
The project implementation will involve not only testing individual algorithms for the operation of various types of construction equipment but also the development of unified data exchange protocols for all machines, as well as a system for telemetry and precise geospatial positioning of equipment on the construction site.
In global practice, a similar project, Concept X, has been implemented since 2022 by a consortium of companies including Hyundai and Doosan, with an expected commercial solution date by 2025. In our opinion, combining the efforts of leading Russian companies could enable the implementation of a similar project within comparable timeframes, given that individual basic technologies have already been developed in Russian companies and are not inferior to foreign counterparts.
Regarding the integration of the Horizon Project into the technological agenda within state programs for the development of various industries: The Horizon Project fully aligns with the Strategy for the Development of Unmanned Aviation, adopted in June 2023. The implementation of this strategy will fully provide the Horizon Project with the necessary technologies for the continuous monitoring of the landscape and working conditions of construction equipment using UAV data.
The Horizon intelligent platform itself, in turn, fully corresponds to the concept of end-to-end digital technologies, adopted as a key support tool within the federal project "Digital Technologies" of the national program "Digital Economy of the Russian Federation."
The project implementation will involve not only testing individual algorithms for the operation of various types of construction equipment but also the development of unified data exchange protocols for all machines, as well as a system for telemetry and precise geospatial positioning of equipment on the construction site.
In global practice, a similar project, Concept X, has been implemented since 2022 by a consortium of companies including Hyundai and Doosan, with an expected commercial solution date by 2025. In our opinion, combining the efforts of leading Russian companies could enable the implementation of a similar project within comparable timeframes, given that individual basic technologies have already been developed in Russian companies and are not inferior to foreign counterparts.
Regarding the integration of the Horizon Project into the technological agenda within state programs for the development of various industries: The Horizon Project fully aligns with the Strategy for the Development of Unmanned Aviation, adopted in June 2023. The implementation of this strategy will fully provide the Horizon Project with the necessary technologies for the continuous monitoring of the landscape and working conditions of construction equipment using UAV data.
The Horizon intelligent platform itself, in turn, fully corresponds to the concept of end-to-end digital technologies, adopted as a key support tool within the federal project "Digital Technologies" of the national program "Digital Economy of the Russian Federation."