Foresty and WoodiBot

Following the sharing of information to over 300 experts in the forest engineering and logistics engineering fields via email and social media, I proceeded to conduct interviews with a total of 10 experts. Among these experts, 4 had Master’s degrees, 4 held Ph.D.s, and 2 were professors. Master’s students were regarded as experts in the two situations. Firstly, they engage in the completion of a master’s thesis. Secondly, they must have a minimum of 2 years of experience in the forest industry. The mean age is 32.8 years, with one female and nine males. Nine of the participants are experts in Forest engineering and 1 of them is from Logistic engineering. The interviews, observations, and usability tests were done in Zoom and Google Meet. The list of participants is shown in Table 1.

The abnormal day scenario describes a challenging day in timber supply transportation in Morice, Canada, featuring a team (our personas) including drivers Michael, Thomas, Marc, Taylor, admin Alex, WoodiBot (a chatbot), station manager Betty, factory manager Hossain, and mechanical engineering manager David. The day starts with Alex discovering absences due to family issues and using WoodiBot to manage logistics. Despite these challenges, including a mechanical breakdown and staffing issues, the team managed, with WoodiBot coordinating driver efforts and managing the collection and transportation of wood, ensuring the day’s objectives were met through effective communication and technology utilization. Our scenario illustrates problem-solving by using human-AI communication and collaboration in an autonomous system in the face of unexpected obstacles, highlighting the importance of flexibility and digital tools in logistics forestry. Below, you can see a part of the communication between Woodibot in the Foresty application and different users.

The normal day scenario involves a smoothly operating team, including drivers like Mikael, admin Alex, and WoodiBot, a chatbot facilitating interactions. The day starts with routine checks and progresses with efficient wood transportation, thanks to effective planning and communication through the Foresty app. WoodiBot plays a key role in optimizing operations, and ensuring targets are met. This scenario highlights the effectiveness of technology and teamwork in daily logistics management.

The figure illustrates the high-level system architecture and interaction flow of the Forestry application, centered around the Woodibot AI system. The diagram shows how different user roles, including administrators, factory managers, station managers, and drivers, communicate with the system through the Forestry app. Woodibot acts as the core intelligent agent, coordinating information exchange, task management, and decision support across stakeholders. Multiple interaction modalities, such as text, voice, and visual inputs, are supported to enable real-time communication during forestry operations. This architecture highlights Woodibot’s role as an intermediary between human users and operational processes, ensuring efficient coordination, monitoring, and reporting throughout the harvesting and transportation workflow.

Participant P4 described the harmful effects of the group-cut method, stating: “There are numerous ways to gather wood in forests, such as the group-cut method, which involves completely removing all of the trees in a given area. Forest experts have determined that this method is harmful to the biodiversity and texture of the forest.” P4 further explained the evolution of harvesting techniques, emphasizing the need for more sustainable approaches. As noted, “Later, more contemporary techniques of wood harvesting were utilized, although they were deemed less successful. This approach harvests wood depending on the tree’s diameter and other factors.”

The strategic placement of harvesting stations was also highlighted by P4, who explained that harvesting estimates are based on parcels ranging from 50 to 100 hectares. According to the participant, “In certain years, they gather wood depending on its overall volume. So it is preferable to position the stations in the parcels.” This reflects a deliberate effort to balance productivity with minimizing ecological disruption.

In addition, P6 introduced a digital tool feature aimed at supporting sustainable forest management: “The application also allows you to specify the quantity of afforestation daily in the region. For example, we may prepare for future years by knowing what areas have been affected this year.” This functionality underscores the importance of long-term planning and foresight in conservation practices.

Logistical challenges related to wood transportation were described by P7, who compared two types of machinery: “Wood is frequently transported using two trucks: Timberjack and Zetor. One of these devices has rubber wheels, while the other uses chain wheels.” The participant further emphasized route optimization, stating: “It is preferable to attempt to develop a shorter path for transporting the timber, and the proposed routes should be placed in areas with less forest cover.”

Seasonality emerged as another critical factor in harvesting practices. P8 noted, “Another key consideration is the duration of exploitation. I believe fall is the optimum period for exploitation.” In contrast, P4 emphasized winter as the most suitable season for harvesting, explaining: “The harvesting season is an essential feature that should be included in the tool, because the finest harvest season is in the winter. Harvesting is limited during seasons when the tree is awake and expanding due to the significant harm inflicted on forest tissue.” These insights highlight the importance of incorporating seasonal awareness into harvesting planning to reduce environmental damage.

P10 provided a broader, international perspective on harvesting strategies and their environmental consequences, noting: “Different nations have different strategies for removing trees… In Germany, they rapidly learned that this approach had several environmental consequences and passed legislation prohibiting this form of harvesting.” The participant also emphasized the role of human and mechanical resources in shaping forest outcomes: “The changes in forest texture are not coming from the forest side… rather, they are the result of human and mechanical capabilities.” Furthermore, P10 highlighted the importance of data in managing harvesting operations, stating: “The manager is responsible for data reporting… while the administrative manager needs data on human concerns, logistics, and the types of machines and their support.”

Beyond operational concerns, participants emphasized the need for forestry tools to adapt to regional and national contexts. P3 underscored the importance of flexibility: “It is feasible to employ tools in various ways depending on the circumstances in that nation, because each country has a unique forest and planting and harvesting processes.” Similarly, P9 stressed regulatory alignment: “When this tool is produced, it should be adjusted to the legislation of that nation.” P8 further noted variations in monitoring requirements across countries, explaining: “This technique may vary from independent monitoring depending on the situation in each nation. As a result, in certain nations, additional monitoring is needed to determine which trees should and should not be felled.” The mention of surveillance technologies, such as CCTV in some regions, illustrates how forestry management tools must accommodate diverse legal, ecological, and technological contexts.

The economic implications of wood harvesting are critical to the long-term sustainability and viability of forestry operations. Participants provided valuable insights into the financial challenges and economic considerations that shape decision-making in the forestry industry. P5 emphasized the importance of precise operational planning, stating: “We should be able to incorporate each machine’s loading time into our calculations. In this situation, the cutting and loading times should be specified according to the machine model, because it allows for more precise cost estimations.” This perspective highlights the need for detailed operational data to improve planning accuracy, budgeting, and cost efficiency.

P6 addressed broader economic pressures within the sector, noting: “Economic challenges are a top focus for many organizations. As a result, it must first be guaranteed that this instrument will ultimately result in significantly greater economic output for the firm. Following that, organizations should focus on environmental concerns.” This view reflects the prioritization of economic sustainability as a foundational requirement for investing in environmentally responsible practices.

P7 further discussed the cost-effectiveness of minimizing forest degradation, explaining: “It is more cost-effective to cause less forest destruction. For example, if we employ a machine that knocks down leaves and ruins the soil’s potential to develop, no new trees will sprout in those regions in the following years. As a consequence, the forest was decimated.” This insight reveals the long-term economic drawbacks of destructive harvesting methods that undermine soil quality and forest regeneration.

Expanding on the costs of restoration, P7 added: “If the forest is gone, its restoration will take a long time and be expensive. For example, the process of tree development involves the beginning growth of annual trees, also known as pre-season trees, followed by semi-pre-season trees and peak species, which take many years.” This statement underscores the significant time and financial investment required for forest recovery, reinforcing the importance of sustainable harvesting practices to avoid irreversible economic losses.

For example, P1 appreciated the app’s ability to coordinate wood harvesting variables effectively but pointed out the need for improved GPS functionality to record characteristics of harvestable trees, “The Foresty app provides excellent coordination [...] However, utilizing GPS, the characteristics of harvestable trees (tree age, health state, etc.) must be recorded.” This feature is essential for ensuring responsible and informed harvesting decisions. Also, P2 suggested enhancing the app with navigational aids similar to Google Maps and emphasized the importance of simplicity, “It would be helpful if you could show the driver the way on the panel, like on Google Maps... The fewer options there are, the better.”

P3 proposed the inclusion of specific wood types and potential uses within the app, “Enter the sort of wood that will be collected into the tool. You may also code for its potential use in many businesses.” They also recommended integrating climate and weather information to assist in planning, “If it is feasible to show climate and weather information inside the application, it will assist the application management in making judgments.” Moreover, P7 appreciated the dashboard’s functionality, “The main dashboard displays the quantity of collected industrial wood in cubic meters and tons for wood chips.” This feature provides a quick overview of productivity and resource utilization.

P4 and P5 discussed customization and advanced features such as altering the app’s logo based on the research region or tree type, and displaying forest development and potential for further harvesting. P4 noted, “The application’s logo should be altered according to the research region [...] ” while P5 highlighted, “If the application includes the ability to display development in various regions of the forest and replace forests, it will become much more intriguing.” Also, P6 focused on the app’s potential for operational management, suggesting the inclusion of charts for daily withdrawals, truck distances, and staff presence, “In the chart area, you may be able to include a daily withdrawal chart in the app... You may even create a chart showing the presence and absence of staff.”

The usability test of the Foresty app highlighted several key aspects of user interaction, including both strengths and areas for improvement, as perceived by various participants. P3 speculated on the app’s impact on operations engineers, suggesting it might reduce their necessity while still retaining a supervisory role, “I believe this will reduce the need for operations engineers. Perhaps they perform a supervisory function.” Also, P6 recommended the installation of closed-circuit cameras for increased monitoring and independent surveillance, “To increase monitoring, closed-circuit cameras may be installed [...] the administrator has enough capacity to conduct independent surveillance.”

P1 praised the app for its potential value to main managers but emphasized the importance of real-world testing to evaluate user reactions and identify any limitations or errors, “The Foresty app can be utilized by the main manager and is quite valuable [...] this tool should be used in the actual world to assess people’s reactions.” Moreover, P2 pointed out the need for a feature to report unexpected route blockages, expressing a preference for human interaction over automated responses, “The program should include an option to report routes that are blocked [...] I’d rather talk to a live person about this than a chatbot.” Also, P3 mentioned the potential integration of technologies like PlantNet, LeafSnap, Blossom, and PictureThis to assist in recording significant species, suggesting an enhancement to the app’s functionality, “Many technologies, such as PlantNet, LeafSnap, Blossom, and PictureThis, have been created to assist people in recording important species [...] This program captures the location and time of species registration.”

P5 discussed the challenge of assigning harvest points in the real world, a task that the app simplifies but is traditionally time-consuming due to the need for manual volume inventory by foresters, “One of the tool’s benefits [...] but in the real world, this task has proven difficult.” P8 criticized the app’s voice communication features as inadequate, stressing the need for more detailed reporting of tree properties for accurate volume and cost estimations, “In my perspective, the part of the stations where contact is done by voice is weak [...] This application should be able to report the chopped trees’ properties.”

P9 and P10 offered insights into the app’s utility for different management roles, with P9 emphasizing the importance of a straightforward, user-friendly interface, “I believe that users in the administrative and executive areas of work need a straightforward and user-friendly solution.” P10 highlighted the distinction between administrative and technical management roles, noting the necessity of on-site presence for technical managers to address operational challenges, “This technology allows the administrative manager to operate freely and remotely, but the technical manager must be present at the wood harvesting site.”