Evaluation of smart farm training for extension officers to support digitalizing era

Smart farming is an agricultural concept based on precision agriculture. It utilizes technology automation supported by big data management, machine learning or artificial intelligence, and the Internet of Things to improve the quality and quantity of agricultural production. In its implementation, extension workers play a vital role as the vanguard of this program escort. Therefore, increasing the capacity of extension workers in smart farming is very necessary. This paper evaluated several aspects of smart farming training for extension workers. This paper uses secondary data from the evaluation form of 30 participants collected from Batangkaluku Agricultural Training Centre. Results showed that male extension officers dominated the training participants (70%). The subjects/training modules can be mastered by participants with scores>3.4 out of 5. The training can increase participants’ comprehension of the material up to 2.75x higher. The satisfaction level was above 4 out of 5. The findings can help the organizing committee in improving related aspects. Increased capacity building of agriculture extension officers by such training was expected to overcome the gap in the information of technology innovation from researchers to farmers.


Introduction
The growth in the world population increases the demand for food production.In addition, the reduction in the labor force in rural areas and the increase in production costs pose challenges for food production nowadays [1].Current agricultural land is constrained by various variables, including land and climatic patterns, population density, and growing urbanization, which pose continual challenges to the supply of arable land.As a result, the gap between food demand and availability grows more prominent and problematic over time [2].
Smart farming is a new farm management concept that employs techniques and technologies of agricultural production to overcome obstacles in food production demands and labor reduction [3].Smart farming uses sensors to collect data (temperature, humidity, light, pressure, presence), communication networks to send and receive data, and MIS and data analysis tools to organize and analyze the data.At the same time, the Internet of Things (IoT) connects networked objects [4].Smart farming relies on quantifiable data from sensors on the land, including land identification, weather/climate, crop identification at each place, soil conditions, fertilizers, seeds, pesticides, harvests, crop damage, production, and marketing amounts.Existing sensors should offer real-time land and crop conditions information to prescribe fertilization, irrigation, pesticides, or harvest schedules.Costs associated with adopting technology for individual farms and a lack of knowledge and skills can be substantial barriers to adoption, particularly in developing nations.Thus, access to cutting-edge technologies may be limited to large, industrialized farms [5].
Training and workshop are among the keys to disseminating smart farming.Extension officers are spearheads in spreading and connecting farmers to technology and innovation.Improving farmers' human resources can be optimized through extension activities if carried out by extension workers who have high capacity and are competent in performing.
Human resources must be appropriately managed by providing more knowledge, skills, and expertise following needs [6].In addition, the performance and motivation of extension workers are influenced by characteristics, competencies, and work environment, which includes infrastructure and relationships between people in it [7].Thus, the capacity building of the extension officers needs to continue to be increased through training programs, including smart farming.This paper aims to evaluate smart farming training activities for extension workers.

Methods
This research uses a cross-sectional approach and employs secondary data.The data was obtained through a paper-based questionnaire from the training evaluation process [8].The total respondents were 30 agricultural extension workers from around Sulawesi participating in Smart Farming Training at Batangkaluku Agricultural Training Center.Data was statistically analyzed using a descriptive statistic, paired t-test for pre-test and post-test parameters, and Pearson's correlation test.

Comprehension level.
The comprehension level was scored per module/subject according to the Likert scale (1)(2)(3)(4)(5).The lowest score was 1, and the highest score was 5.It was a self-assessment that each participant was requested to fill the form.Score interpretation was described below: >4.4 very mastery; >3.4 master; >2.4 mastering enough, >1.4 lack of master, <1.4 no master.

Evaluation of the learning process.
Evaluation of the learning process was the average pre-test score and post-test score.The pre-test score was undertaken before the training begins.The post-test was held after all modules are given (before the closing).The exam and marking are the trainer's responsibility for each module.The range of the marking is 0(lowest)-100 (highest).The final score also ranged from 0-100, the detail of interpretation as follows:<=100 Satisfying;<=90 Very good;<=80 Good; <=70 Enough;<=60 Less.

Evaluation of satisfaction level.
Scoring was based on participants' perspectives.The range of score was from 1 (lowest) to 5 (highest).The parameter of observation consists of administration service and training facility service.

Distribution of participants' origin and gender.
The targeted Smart Farming Agriculture training participants were agriculture extension officers from Sulawesi area.Most participants were from South Sulawesi (15), and 3 from each province in Sulawesi (Figure 1).Participation in the training is the authority of the official who was addressed (local service).Generally, it was in the form of a direct appointment from the head office.Participant's gender was dominated by males, who contributed for 70% (21 person), and only 9 females or 30% contribution.If any correlation between gender and some evaluation aspect will be discussed below in a related subtopic (Figure 2).

Comprehension level on each subject.
Eight modules were given during the training from 21-28 March 2022 (Figure 3).Modules were delivered according to the syllabus from the trainers.In general, the modules/subjects can be mastered by participants with an average score of the mastering category (>3.4).The range of scores was from 3.47 (Module Programming and Assembly for Smart Farming) to -4.20 (People's Business Credit (KUR)).
The subject of Module Programming and Assembly for Smart Farming is essential in smart farming.Thus, the low score in that subject required more attention from the committee and the trainer to evaluate the cause and find the solution.It was suggested that the participants of smart farm training have a basic background or experience in computation, electricity, or engineering.The average score of mastery of the subject by participants for all materials is 3.96 (in the mastery category).No significant correlation was found between gender and participants' understanding of the material ( corr -0.001; Sig.(2-tailed) 0.996).

Evaluation of the learning process.
Based on Figure 4, there was an increase in participants' knowledge after attending smart farming training.The questions on pre-test questions include all the material to be delivered.This pre-test determines respondents' knowledge level before delivering the material and knowledge.The pre-test value obtained by participants will encourage interest and increase the enthusiasm of participants to play an active role during the training.Statistical analysis of paired t-test samples showed a significant difference between pre-test and posttest scores.The average score value is 29±SD12.69,while for the post-test, the average is 84±SD 16.31.The average increase in understanding after training of 275.4% means that the training can increase participants' understanding/mastery of the material up to 2.75x higher than before.Pearson's correlation test between genders and increased comprehension showed no significant correlation (corr 0.063; Sig. (2-tailed) 0.740).
According to [10], the provision of pre-test and post-test is one of the teaching methods to improve student learning outcomes in teaching and learning activities.Furthermore, giving pre-tests and posttests can help teachers to evaluate and improve activities and teaching methods.One of the indicators of successful training is the increase in trainees' knowledge [11].

Participant's satisfaction with training implementation.
The satisfaction score ranged from 4.13-4.67from the 5.00 maximum score (Figure 5).The highest satisfaction aspects were in the administration (fast, easy, and friendly registration process) and officers' professionalism and hospitality.The lowest satisfaction score was the hospitality and tidiness of the dining room attendants.No significant correlation was found between gender and satisfaction score.
This finding can guide the committee to evaluate the activities in the dining room.The quality of training materials (accessories) was also the second lowest.These accessories include a bag, T-shirt, and basic stationery.High satisfaction with the implementation of training is influenced by the effectiveness of training and the quality of services provided during training [12].Another study also mentioned that trainees would be motivated if the training process goes well [13]

Conclusion
Based on the evaluation of discussed factors, the Smart Farming Training for extension officers has been successful, as indicated by the high scores in comprehension for each module, the significant increase in post-test scores compared to pre-test scores, and the high satisfaction levels.However, a successful arrangement for Smart Farming Training must be followed by proper implementation after the training to prepare for the digital era of agriculture.

Figure 1 .Figure 2 .
Figure 1.Data distribution on the origin of smart farming trainees

Figure 3 .
Figure 3. Trainee's comprehension level of each subject

Figure 4 .
Figure 4. Evaluation of learning process based on pre and post-test

Figure 5 .
Figure 5. Level of participant satisfaction with the implementation of training

3,80 3 ,
90 4,00 4,10 4,20 4,30 4,40 4,50 4,60 4,70 4,80 Fast, easy and friendly registration process Qualified training materials Officer's professionalism and hospitality Easy and fast payment of settlement Professionalism and hospitality of finance staff Convenience of the dormitory Convenience of the study room Convenience of the dining room Hospitality and tidiness of the accommodation staff Hospitality and tidiness of the dining room attendants Availability of teaching aids (LCD, OHP, Projector,… Committee Teaching Accommodation and consumption