Feedback, please! New research enables autonomous forestry machines to communicate intelligently

The autonomous forestry machines of tomorrow will certainly be able to manage a great deal on their own, but their supervisors still need regular updates on how the work is progressing. Mistra Digital Forest has identified the type of feedback operators want to receive – these insights are currently being used in a prototype for autonomous site preparation and planting, due to be ready in 2026.

Site preparation and planting work can be physically demanding. Spending many hours in the cab of a bumpy machine exposes the body to vibrations, and the outcome is largely influenced by the operator’s prior experience. Several research projects are developing autonomous solutions that could improve the working environment in the forest, and make the end result more predictable. However, for these innovations to be of use, the machine and the operator need to be able to communicate effectively. A project within Mistra Digital Forest has been investigating exactly what this entails.

– It’s easy to create something brilliant ‘on paper’, but it’s quite another matter to develop machines that are actually used as intended. If new technology is to be useful, we have to start with identifying user needs. What makes us want to use new technology, and what kind of information do we need in order to monitor the work of an autonomous machine? Our project´s aim is to identify what type of feedback the machine needs to provide in order to be relevant to its operator, says Madeleine Silverbratt at Skogforsk. Together with SLU, she has been leading the project within Mistra Digital Forest.

Three key areas have been identified 

As this type of autonomous machine is not yet available on the market, the project has assessed the usability of existing systems. Machine operators with varying levels of experience were interviewed in order to find out what kind of feedback they would like to receive. Based on these insights, the project then analysed what the future requirements might look like.

– We have identified three key areas that need to be addressed. The machine needs to alert the operator if technical faults occur, and to indicate how the machine settings should be adjusted to suit the site conditions. The user also needs to receive information about the results of the work. If the quality of the work deviates from expectations, the operator needs to be informed promptly that something has happened, and what action needs to be taken, says Madeleine Silverbratt.

Justified investments according to the forestry contractor 

Among other things, the results of the study will be used in the development of Autoplant, a system for autonomous site preparation and planting, which is being developed in collaboration between the research sector and the industry. A prototype of this new automatic site preparation and planting device is scheduled to be ready in 2026. It is one of several ongoing development projects within forestry automation.

– The development of autonomous and remote-controlled solutions has progressed surprisingly well, and we hope to be able to conduct tests out in the field soon. My view is that these are justified investments and that operators can transition relatively easily to certain semi-automated solutions. Manual handling will always be required to some extent, so we need machines that are able to communicate effectively with the operator in their role as supervisor, problem-solver and decision-maker, says Magnus Thor from forestry contractor Magnus Thor.

Semi-autonomous solutions to be operational within a few years 

The machine manufacturer Bracke Forest is the company that is developing the Autoplant prototype, with the aim of bringing a product to market within a couple of years. They believe that the findings from Mistra Digital Forest’s study can also be applied to the development of existing systems. For example, these insights could improve user interfaces, and streamline information flows.

– We try to identify the needs of the operator at an early stage, so that both the interface and the feedback within the system are designed in a way that works in practice. It is not the technological development itself that often takes time, but rather the implementation and verification that the systems operate in a stable manner, and that users feel confident when using them. Given the pace of development that we are currently seeing, I believe we will have fully functional, semi-autonomous solutions in operation within a few years, and some fully autonomous systems in forestry a few years after that, says Ruben van Westendorp, R&D Manager at Bracke Forest.