Comparison of Mechanized and Automated Technologies in the Scope of Cumulative Energy in Sustainable Milk Production

Abstract

The labor and energy inputs for milking and milk cooling, the preparation and distribution of feed, and manure removal were analyzed on eight farms in the Masovian, Podlaskie, Lower Silesian, and Kuyavian–Pomeranian voivodeships. In these facilities, milking was carried out using milking robots, which were also employed during feed preparation and discharge. The cleaning of manure channels and walking corridors were also automated. Based on labor, fuel, energy consumption, weight of machinery and equipment, and building area, the cumulative energy intensity of the farms was calculated, taking into account the average density of cow herds. The lowest energy input, 0.974 kWh∙LU⁻¹∙day, and the lowest cumulative energy intensity, amounting to 19.601 MJ∙day⁻¹∙LU⁻¹ were observed in a farm with AF-AM5 technology, that is, an automated milking system (AMS) and an automated feeding system (AFS). The second lowest total energy intensity was measured in a farm utilizing the AF-AM7 technology and amounted to 22.505 MJ∙day⁻¹∙LU⁻¹. The highest cumulative energy intensity was measured in a farm practicing conventional milking in a fishbone milking parlor, with a traditional feeding wagon powered by an agricultural tractor. A high level of cumulative energy intensity, for the majority, was connected with a high energy and fuel consumption, especially during the preparation of feed and feed distribution.

1. Introduction

One of the goals of the Common Agricultural Policy 2022–2027 is to reduce greenhouse gas (GHG) emissions and improve animal welfare [1]. In line with the European Green Deal, in the “From Farm to Fork” strategy, agriculture should increasingly adapt to climate change, and actions should be taken to neutralize its negative impact [2]. A reduction in greenhouse gas emissions can be achieved, among other solutions, by using innovative technical solutions that save energy from conventional sources [3,4,5]. Sustainable development in agriculture includes, among other aims, minimizing agriculture’s negative impact on the environment [5].

The development of modern livestock breeding technologies should be associated with the rational storage and management of fertilizer through methane fermentation, allowing for the production of energy in the form of biogas [6,7] or other resources [8,9].

World leaders in milk production have introduced solutions that automate activities in the processes of milking, preparation and distribution of feed, and removal of natural fertilizers. Scientific centers dealing with the issue of energy and labor consumption in animal production include the following: Technische Universität München; Hochschule Weihenstephan-Triesdorf; College of Agriculture, Food and Natural Resources, at the University of Missouri; Agroscope; and Kuratorium für Technik und Bauwesen in der Landwirtschaft (KTBL).

On family dairy farms in Germany, Poland, and other European countries, as well as around the world, the role of automation in the milk production process is increasing [10]. In light of the shortage of hired labor resulting from two reasons—low wages and a high burden—the automation of milking and/or feeding is becoming more and more popular in Poland. On the one hand, thanks to automation, it is possible to increase the amount of milk obtained (by increasing the frequency of milking and feeding cows more frequently) [11], but compared to the traditional method of milking or feeding, the total energy demand in such automated systems increases [12,13].

The workload in milk production operations is divided into physical, mental, and sensorimotor work. Automation reduces the amount of physical work [14], and the nature of the remaining work changes towards mental work as a result. Milk producers often decide to robotize all technological procedures in the production process or robotize only selected procedures, an option which is most often due to limited investment capacity [15,16].

Milking robots, as the most modern automatic devices in milk production, introduced in Poland in the last fifteen years in several dozen free-stall barns, work without the presence of a human, performing all milking activities as follows: they recognize a cow using various types of sensors (ultrasound, laser, and camera) [17,18], provide a massage of the teats, put on the teat cups, remove them in the order in which the quarters are milked, wash the teat cups after each cow, and carry out hygienic spraying of the udders after milking [19]. The use of milking robots facilitates an increase in the frequency of milking cows, reducing labor intensity by 20% or even 50%, and increasing milking efficiency by 5% or even 25% [20,21,22]. Schick claims that the use of milking robots, compared to the traditional milking system, leads to a significant and even two-fold reduction in the labor intensity concerning milking procedure [23].

Similarly, according to Heitkämper et al. [24], the use of milking robots, compared to the traditional milking system, also leads to a significant reduction in labor intensity. Significantly lower human labor inputs in barns with AMS were obtained in the research by Poulopoulou et al. [25].

Due to increasing interest in investing in robotized systems for milk production, there is a lack of precise data about the energy efficiency of said systems. The aim of this study was to investigate the impact of the machines and devices used on the work and energy inputs in cattle production buildings as well as the cumulative energy inputs in order to indicate the best technology model with the lowest cumulative energy needs. The scope of this research included the assessment of the electrical and mechanical energy consumption required to produce 1 L of milk from eight herds of dairy cows ranging in size from 83 LU to 320 LU and with an average annual milk yield ranging from 7500 L to 11,400 L.

2. Materials and Methods

The subjects of the analysis in this article were the labor and energy expenditure concerning the activities of milking, milk pre-treatment, preparation and distribution of feed, and removal of natural fertilizers in robotic barns.

Eight barns characterized by high level-IV mechanization were selected for the study, i.e., the daily unit of labor input was less than five labor-minutes per 1 day and per livestock unit.

A description of the tested farms is presented in

Table 1. Main characteristics of the tested farms, including cattle barns.

Technology	Number of LU (Milking Cows)	Housing System	Milking and Milk Cooling	Annual Milk Yield [dm3∙LU−1]	Building Area [m2]
CF-AM1	160	Boxed, free-stall, deep slurry channels	2 milking robots, producer 1	8500	2113
CF-AM2	83	Boxed, free-stall, deep slurry channels	1 milking robot, producer 2	7500	1298
CF-AM3	200	Boxed, free-stall, deep slurry channels	2 milking robots, producer 3	11,000	1543
CF-AM4	200	Boxed, free-stall, deep slurry channels	2 milking robots, producer 3	9300	4160
AF-AM5	170	Boxed, free-stall, deep slurry channels	3 milking robots, producer 3	10,500	1800
CF-AM6	230	Boxed, free-stall, deep slurry channels	4 milking robots, producer 1	11,903	2220
AF-AM7	320	Boxed, free-stall, deep slurry channels	4 milking robots, producer 1	11,400	3001
CF-CM	200	Boxed, free-stall, shallow litter	Fishbone2x8, producer 2	9500	2194

CF-AM1, CF-AM2, CF-AM4, CF-AM6—conventional feeding (CF) and robotic milking (AM); AF-AM5 and AF-AM7—automated feeding and robotic milking; CF-CM—conventional feeding and conventional milking.

, including the types of technology tested, the herd size, the average annual unitary milk yield, and the total building area.

Seven barns used a bedding-free box housing system for the cows, with deep manure channels. In one (AF-AM7) there were shallow manure channels and in another (CF-CM) a box bedding system. In the 7 barns with deep channels, the slotted floor was further cleaned by a robot. Robots for the slotted floors moved along the floors with slurry channels underneath, and pushed animal excrement into them. Such a robot avoids obstacles, including cow legs. It uses remote control, and it is possible to plan a route and enter it into the robot’s memory. In 1 barn (CF-CM), the slurry was scraped from the manure aisle using a “delta” scraper.

Milking in 7 barns was performed using milking robots (CF-AM1 to AF-AM7), while only 1 barn (CF-CM) used conventional milking with a fishbone milking parlor where 2 × 8 stalls were used. The number of milking robots differed depending on the herd size, but was from 1 to 3 milking units.

Table 2. Characteristics of the machinery and equipment used for the following treatments: T_II—feeding and feed preparation; T_III —manure removal and storage.

Technology	Treatment TII			Treatment TIII
CF-AM1	Feeding wagon, tractor	Concentrates in milking robots	Robotic feed-scraper	Slurry tank, tractor
CF-AM2	Feeding wagon, tractor	Concentrates in milking robots	Robotic feed-scraper	Slurry tank, tractor
CF-AM3	Feeding wagon, tractor	Concentrates in milking robots	Robotic feed-scraper	Slurry tank, tractor
CF-AM4	Feeding wagon, tractor	Concentrates in milking robots	Robotic feed-scraper	Slurry tank, tractor
AF-AM5	TMR robot (fully automated), producer 3	Concentrates in feeding robot	Feed pushed by PMR robot	Slurry tank, tractor
CF-AM6	Feeding wagon, tractor. Self-power loader	Concentrates in milking robots	Robotic feed-scraper	Slurry tank, tractor
AF-AM7	TMR robot. producer 2	Concentrates in milking robots	Feed pushed by TMR robot	Slurry tank, tractor
CF-CM	Self-power-feeding wagon	Robot used for concentrates	Robotic feed-scraper	Slurry tank, tractor

CF-AM1, CF-AM2, CF-AM4, CF-AM6—conventional feeding (CF) and automated milking (AM); AF-AM5 and AF-AM7—automated feeding and automated milking; CF-CM—conventional feeding and conventional milking.

shows the details for the feeding systems used in the barns.

In 6 barns, a pushing robot was used to push feed on the feed table. Only in two barns, AF-AM5 and AF-AM7, was the feed dispensing system fully automated, while in the remaining facilities, the PMR feed mixture was fed using a feed wagon.

In the case of automated milking, at least one milking box with one unit is located in the non-central part of the barn. In addition, robots automatically measure the milking volume and assess the milk quality, dispensing a portion of granulated concentrate feed. The milking robot was integrated electronically as part of a computer system for cow herd management [20]. In 6 of the tested barns with milking robots, automatization was organized for the feeding process, based on Total Mixed Ratio (TMR) feeding systems. Supplementary concentrate feed was provided at the milking robot station. The two fully automated barns (AF-AM5 and AF-AM7) were equipped with TMR robots, computer-controlled machines that collect feed components and, after mixing it, deliver feed to the feed aisle [26].

After programming the groups of animals in the computer (number of animals in a group, amount of feed per animal, feed mixtures, the share of individual components in the feed, corn, haylage, and food concentrate) and the feeding times, the TMR robot automatically controls the entire feeding process [27].

Feeding robots for concentrates and/or roughage may be used in stall barns and consist of independent dispensers and other equipment to identify the cow positions [28]. The dispenser works completely automatically, dispensing the feed according to a programmed daily feeding cycle, where the involvement of the service is limited to periodic changes in data in the controller memory along with inspections and repairs [29]. In a free-stall system, the automatic system for feeding cows with concentrates consists of a feeding machine and an electronic controller that identifies the cow anywhere on the farm. Such solutions have been used in two cattle barns tested with automated feeding and automated milking (AF-AM5 and AF-AM7).

In all eight barns, ventilation was natural, supported by air mixers, while electric dehornizers were used in the IV treatment. In the AF-AM5 barn, there was a system of air curtains, creating a barrier to the external environment.

In the matter of automation of other systems in all the locations, a cattle brush was used to meet the behavioral needs of the animals related to the given environment. There are two types of cattle brushes available: non-powered brushes, and electric brushes that include a motion sensor and an electric motor. The second type was used in all the locations.

The technological characteristics of all the facilities are presented here, along with the concentrations of the herds. The barns were characterized in terms of actual labor as well as the electrical and mechanical energy inputs. The daily labor inputs were determined using the timing of the working day. Machines and other equipment were identified for the mechanization of individual technological procedures constituting the raw material production processes, including the consumption of diesel fuel. Measurements of electrical consumption by the machines and other equipment were carried out using a porable 3-phase Elite Logger Pro recorder.

Cumulative energy consumption was calculated based on the technological data collected in a tabular form and the conversion factors (

Table 3. Conversion indicators of cumulative energy consumption.

Indicator	Indicator Value
Human work (L)	40.00 MJ/rbh
Electrical energy (E)	13.60 MJ/k/Wh
Diesel fuel (F)	53.20 MJ/kg
Buildings area (B)	100.00 MJ/m2
Machinery and other equipment (W)	110.00 MJ/kg

Definitions: human work (L), electrical energy (E), diesel fuel (F), machinery and other equipment (W).

) provided by Wójcicki [30].

Cumulative energy intensity was calculated as a product of particular items: labor, fuel, electrical energy, building area, and equipment weight multiplied by the indicator values. Cumulative energy intensity was the sum of the energy intensity resulting from the labor inputs, electrical energy, fuel, building area, and weight of the machinery and other equipment related to mechanization and robotization, as in Equation (1).

E_C = E_CL + E_CF + E_CE + E_CM + E_CB

(1)

where the following are defined:

• E_C—cumulative energy intensity [MJ∙day⁻¹∙LU⁻¹];
• E_CL—cumulative energy intensity for labor, [MJ∙day⁻¹∙LU⁻¹];
• E_CF—cumulative energy intensity for fuel, [MJ∙day⁻¹∙LU⁻¹];
• E_CE—cumulative energy intensity for electrical and mechanical energy, [MJ∙day⁻¹∙LU⁻¹];
• E_CM—cumulative energy intensity for machinery, connected to weight, [MJ∙day⁻¹∙LU⁻¹];
• E_CB—cumulative energy intensity for buildings, [MJ∙day⁻¹∙LU⁻¹];
• LU—livestock unit.

Statistical Analysis

In order to check whether there were any correlation between herd size and cumulative energy factors, a Pearson correlation analysis was performed using the Python language in a programing environment of a Jupiter Notebook. The results of the statistical analyses were tabularized and are presented in the Figures.

Equation (2) was used as follows:

$$r_{XY}=\frac{co{v}_{XY}}{{\sigma }_X{\sigma }_Y}=\frac{\sum \left(X-\overline{X}\right)\left(Y-\overline{Y}\right)}{\sqrt{\sum {\left(X-\overline{X}\right)}^2\sum {\left(Y-\overline{Y}\right)}^2}}$$

(2)

where r is the Pearson correlation, X is LU (livestock Unit), and Y is the cumulative energy factors, respectively (for each technological treatment and/or cumulative energy sources: labor (E_CL), fuel (Ec_F), electrical and mechanical energy (E_CE), weight of machinery and equipment (E_CW), and building area (E_CB)).

3. Results

3.1. Technological Parameters Characterizing the Objects Tested

Table 4. Total weight of machinery (kg) and equipment for particular technological treatments: T_I—milk and milk cooling; T_II—feeding; T_III—manure removal and storage.

Technology	TI	TII	TIII	Total Weight W(I+II+III)
CF-AM1	6340	10,370	26,608	43,318
CF-AM2	3170	15,935	10,920	30,025
CF-AM3	7240	15,780	23,814	46,834
CF-AM4	9180	14,083	15,300	38,563
AF-AM5	5540	11,075	16,778	33,393
CF-AM6	7970	20,965	11,495	40,430
AF-AM7	9180	10,725	18,435	38,340
CF-CM	3190	12,558	11,660	27,408

CF-AM1, CF-AM2, CF-AM4, CF-AM6—conventional feeding (CF) and robotic milking (AM); AF-AM5 and AF-AM7—automated feeding and robotic milking; CF-CM—conventional feeding and conventional milking.

shows the total weights (kg) of all machinery and equipment used in the mechanization and automatization of technological treatments T_I, T_II, and T_III. As the weight of the equipment for treatment T_IV (other procedures) was relatively small, it was decided to remove this item from further analysis.

The heaviest machinery and equipment were to be found in the cattle barn using a conventional feeding system, equipped with a tractor and feeding wagon along with milking by using two milking robots (CF-AM3). The lowest weight of the machinery was in the location with conventional feeding and milking (CF-CM).

In the conventional feeding technologies (CF) in treatment II (T_II), the consumption of diesel fuel, presented in

Table 5. Daily fuel consumption (F) by tractors used in the T_II and T_III treatments [kg∙day⁻¹∙LU⁻¹].

Technology	FTII	FTIII	Total F
CF-AM1	0.059	0.047	0.106
CF-AM2	0.066	0.023	0.089
CF-AM3	0.098	0.032	0.130
CF-AM4	0.064	0.017	0.081
AF-AM5	0.022	0.002	0.024
CF-AM6	0.071	0.005	0.076
AF-AM7	0.042	0.007	0.049
CF-CM	0.105	0.382	0.487

CF-AM1, CF-AM2, CF-AM4, CF-AM6—conventional feeding (CF) and robotic milking (AM); AF-AM5 and AF-AM7—automated feeding and robotic milking; CF-CM—conventional feeding and conventional milking.

, resulted from the operation of an agricultural tractor with a mixing feed wagon and a loader for loading feed. In the robotic technology systems it was the consumption generated by the tractor for loading roughage into the feed hoppers as part of the “forage cuisine”. In the T_III technological procedure for removing natural fertilizers, this consumption resulted from the use of tractors to pump out the liquid natural fertilizer in the form of slurry.

The highest daily unitary diesel consumption (0.487 kg∙day⁻¹∙LU⁻¹) was observed in the conventionally mechanized cattle barn with a self-powered feeding wagon and conventional milking in the fishbone milking parlor (CF-CM). The high diesel consumption may be the result of the high weight of the machinery used for conventional feeding, because it depends on the power which needs to be supplied. In this case, a self-powered wagon was used, which is related to higher fuel needs.

Table 6. Daily unitary labor inputs L for technological treatments: I (milk and milk cooling), II (feeding), III (manure removal and storage), and IV (other procedures).

Technology	LI	LII	LIII	LIV	Total A	Total B *
CF-AM1	0.330	0.350	0.030	0.070	0.780	0.013
CF-AM2	1.080	0.600	0.070	0.170	1.920	0.032
CF-AM3	0.920	1.710	0.090	0.270	2.990	0.049
CF-AM4	0.570	0.430	0.001	0.100	1.101	0.018
AF-AM5	0.441	0.088	0.021	0.099	0.838	0.014
CF-AM6	0.391	0.478	0.022	0.081	1.701	0.028
AF-AM7	1.080	0.600	0.160	0.180	2.020	0.033
CF-CM	2.040	0.710	0.350	0.310	3.410	0.057

* A—labor-minutes∙day⁻¹∙LU⁻¹; B—labor-hours∙day⁻¹∙LU⁻¹. CF-AM1, CF-AM2, CF-AM4, CF-AM6—conventional feeding (CF) and robotic milking (AM); AF-AM5 and AF-AM7—automated feeding and robotic milking; CF-CM—conventional feeding and conventional milking.

shows the results of tests on daily unitary labor inputs (L). The lowest total labor inputs (0.018 labor-hours∙day⁻¹∙LU⁻¹) were obtained in the most automatized barn, with automated feeding and automated milking (AF-AM). In contrast, as was expected, the highest labor needs were observed in the conventionally equipped cattle barn marked as CF-CM, at 0.057 labor-hours∙day⁻¹∙LU⁻¹.

Table 7. Daily unitary electrical and mechanical energy inputs for production treatments [kWh∙day⁻¹∙LU⁻¹] (E_I—milking and milk cooling; E_II—feeding; E_III—manure removal and storage; E_IV—other procedures).

Technology	EI	EII	EIII	EIV	Total
CF-AM1	1.207	0.459	0.005	0.167	1.838
CF-AM2	0.781	0.789	0.250	0.182	2.002
CF-AM3	0.735	1.848	0.605	0.121	3.309
CF-AM4	1.232	0.840	0.207	0.228	2.507
AF-AM5	0.629	0.237	0.020	0.089	0.974
CF-AM6	0.698	0.621	0.034	0.135	1.487
AF-AM7	0.648	0.446	0.022	0.111	1.227
CF-CM	0.191	1.826	0.157	0.217	2.391

CF-AM1, CF-AM2, CF-AM4, CF-AM6—conventional feeding (CF) and robotic milking (AM); AF-AM5 and AF-AM7—automated feeding and robotic milking; CF-CM—conventional feeding and conventional milking.

presents the electrical and mechanical energy inputs for particular technological treatments in the milk production procedure. The largest energy inputs were stated for treatment I, i.e., milking and milk cooling, and it was true for each object being tested. The lowest energy consumption for milking was related to conventional feeding and milking (CF-CM), meaning delivering the feed by using a tractor equipped with a mixing wagon and performing the milking procedure in a fish-bone milking parlor. Taking into account the whole technology process, which consisted of four main treatment procedures (milking, feeding, manure removing, and other procedures), the best results, namely the lowest energy supply, were needed in the fully automated (both feeding and milking) cattle barn, while the highest energy consumption occurred in the third barn with automated milking (CF-AM3) and the barn with conventional feeding and milking (CF-CM), where human labor still had a significant role.

Table 8. Cumulative daily energy inputs for treatment T_I—milking and milk cooling [MJ∙day⁻¹∙LU⁻¹].

Technology	ECL	ECE	ECW	Total ECI
CF-AM1	0.220	0.562	16.415	17.197
CF-AM2	0.720	0.576	10.622	11.918
CF-AM3	0.613	0.545	9.996	11.154
CF-AM4	0.380	0.988	16.755	18.123
AF-AM5	0.294	0.726	8.554	9.574
CF-AM6	0.261	0.858	9.493	10.612
AF-AM7	0.720	0.432	8.813	9.965
CF-CM	1.360	0.240	2.597	4.197

CF-AM1, CF-AM2, CF-AM4, CF-AM6—conventional feeding (CF) and robotic milking (AM); AF-AM5 and AF-AM7—automated feeding and robotic milking; CF-CM—conventional feeding and conventional milking; E_CL, E_CE, E_CW—cumulative energy intensity in the scope of labor, electrical energy, and weight of machinery.

,

Table 9. Cumulative daily energy inputs E_CII for treatment T_II—feeding and preparation of feed [MJ∙day⁻¹∙LU⁻¹].

Technology	ECL	ECF	ECW	EcE	Total ECII
CF-AM1	0.233	3.139	0.919	6.242	10.534
CF-AM2	0.400	3.511	2.893	10.730	17.535
CF-AM3	1.140	5.214	1.189	25.133	32.675
CF-AM4	0.287	3.405	1.516	11.424	16.631
AF-AM5	0.059	1.170	1.451	3.223	5.903
CF-AM6	0.319	3.777	2.257	8.446	14.798
AF-AM7	0.400	2.234	0.505	6.066	9.205
CF-CM	0.473	5.586	0.946	24.834	31.839

CF-AM1, CF-AM2, CF-AM4, CF-AM6—conventional feeding (CF) and robotic milking (AM); AF-AM5 and AF-AM7—automated feeding and robotic milking; CF-CM—conventional feeding and conventional milking; E_CL, E_CF, E_Cw, E_CE—cumulative energy intensity in the scope of labor, fuel, weight of machinery, electrical and mechanical energy.

and

Table 10. Cumulative daily energy inputs E_{CIII, IV} for the following treatments: T_III—manure storage and removal; T_IV—other procedures [MJ∙day⁻¹∙LU⁻¹].

Technology	TIII				TIV		Total
	ECL	ECF	EcW	EcE	ECL	EcE	ECIII, ECIV
CF-AM1	0.020	2.500	2.358	0.068	0.047	2.271	7.265
CF-AM2	0.047	1.224	1.983	3.400	0.113	2.475	9.241
CF-AM3	0.060	1.702	1.794	8.228	0.180	1.646	13.611
CF-AM4	0.001	0.904	1.647	2.815	0.067	3.101	8.535
AF-AM5	0.014	0.106	2.198	0.272	0.066	1.210	3.867
CF-AM6	0.015	0.266	1.237	0.462	0.054	1.836	3.870
AF-AM7	0.107	0.372	0.868	0.299	0.120	1.510	3.276
CF-CM	0.233	20.322	0.878	2.135	0.207	2.951	26.727

CF-AM1, CF-AM2, CF-AM4, CF-AM6—conventional feeding (CF) and robotic milking (AM); AF-AM5 and AF-AM7—automated feeding and robotic milking; CF-CM—conventional feeding and conventional milking; E_CL, E_CF, E_CW, E_CE—cumulative energy intensity in the scope of labor, fuel, electrical and mechanical energy ECE, weight of machinery and equipment, and building area.

present the cumulative energy intensity in the scope of particular technological treatments: milking (I), feeding (II), manure removal (III), and other procedures (IV). Regarding milking and milk cooling (Treatment II), the highest energy inputs were registered, while the highest values of cumulative energy were obtained in the automated milking barn CF-AM4 (Table 8).

In general, among all the technological treatments, feed preparation and feeding (Treatment II) was the most cumulatively energy-consuming process (Table 9). The highest cumulative energy intensity was achieved in barns equipped with conventional machinery for feeding (barns CF-CM and CF-AM3).

In treatment T_III—manure removal and storage—similarly to treatment T_II, the highest cumulative energy inputs were obtained from the barn with conventional feeding, milking, and manure removal system by manure scraper, as shown in Table 10.

Figure 1 shows the cumulative energy inputs, divided into particular technological treatments for the different technologies tested. The lowest cumulative energy inputs were observed in two farms (AF-AM5 and AF-AM7) with totally automated milking and feeding, amounting to 19.344 and 22.446 MJ∙day⁻¹∙LU⁻¹. The highest (62.763 MJ∙day⁻¹∙LU⁻¹) occurred in one barn featuring totally conventional milking and feeding (CF-CM). In the group of farms with automated milking and with conventional feeding (CF-AM1 to CF-AM7), the average cumulative energy amounted to 40.739 MJ∙day⁻¹∙LU⁻¹ with a standard deviation of 10.652 MJ∙day⁻¹∙LU⁻¹. The highest share of cumulative energy (E_C) was observed in treatment II—feeding and preparation of feed. Further, in terms of technology, CF-CM had the most energy-consuming treatments among all the barns tested, namely treatment III—manure removal and storage.

Figure 2 presents the cumulative energy intensity in the scope of particular items, including labor, energy fuel, weight of equipment, and building area.

3.2. Results of the Statistical Analyses

The main aim of the statistical analyses was to obtain knowledge about the relations between herd size, labor inputs, and the particular elements of cumulative energy, including in the scope of technological treatments. It could help to formulate the recommendations required for farmers in the area of best sustainable milk production technologies. The correlations were calculated according to Equation (2).

In general, very low Pearson correlation indicators were stated for all the correlations measured, both labor inputs and cumulative energy, in relation to the number of milking cows in particular locations tested with different milking and feeding technologies. The lowest correlation indicators were obtained in the relation between the number of livestock units (LU) versus daily unitary labor, where

Table 11. Pearson correlation factors between labor inputs (L_I, L_II, L_III, L_IV) on technological treatments (I—milking and milk cooling; II—feeding; III—manure removal and storage; IV—other procedures), as well as number of cows (LU).

	Number of Cows (LU)	LI	LII	LIII	LIV
Number of cows (LU)	1.000000	−0.044422	−0.001569	0.034495	−0.057560
LI	−0.044422	1.000000	0.341633	0.949361	0.888941
LII	−0.001569	0.341633	1.000000	0.262262	0.702025
LIII	0.034495	0.949361	0.262262	1.000000	0.829714
LIV	−0.057560	0.888941	0.702025	0.829714	1.000000
EI	0.130708	−0.740841	−0.186054	−0.783278	−0.692538
EII	−0.052151	0.684515	0.793789	0.613860	0.867291
EIII	−0.132317	0.247463	0.899357	0.067359	0.597685
EIV	0.039374	0.383204	−0.060753	0.272763	0.159955

presents the Pearson correlation factors for these results.

Continuing,

Table 12. Pearson correlation factors between cumulative energy (E_CI, E_CII, E_CIII, and E_CIV) on inputs on technological treatments I, II, III, and IV vs. number of cows (LU).

	Number of Cows (LU)	ECI	ECII	ECIII	ECIV	Total
Number of cows (LU)	1.000000	−0.206066	−0.098138	−0.153197	−0.302002	−0.207382
ECI	−0.206066	1.000000	−0.382094	−0.547886	0.151779	−0.212472
ECII	−0.098138	−0.382094	1.000000	0.850053	0.465418	0.954650
ECIII	−0.153197	−0.547886	0.850053	1.000000	0.576802	0.887163
ECIV	−0.302002	0.151779	0.465418	0.576802	1.000000	0.658750
Total	−0.207382	−0.212472	0.954650	0.887163	0.658750	1.000000

L_I—labor inputs in treatment I (milking and milking cooling); L_II—labor inputs in treatment II; L_III—labor inputs in treatment III; L_III—labor inputs in treatment IV; E_I—energy in treatment I; E_I—total energy in treatment II (feeding); E_III—energy in treatment III; E_IV—energy in treatment IV.

gives the correlation factors between herd size (number of livestock units, LU) and cumulative energy for all the technological treatments: E_CI, E_CII, E_CIII, and E_CIV.

In order to know whether the herd size correlates with the sum of cumulative energy in the context of sources, a number of supplementary analyses were undertaken related to the dependences between the number of cows and particular cumulative energy sources (weight of machinery, labor, electrical energy, fuel, and building area), with

Table 13. Pearson correlation factors between cumulative energy and number of cows (LU).

	Number of Cows (LU)	ECL	ECF	ECE	ECW	ECB	Total EC
Number of cows (LU)	1.000000	0.166030	−0.034603	−0.203042			−0.207774
ECL	1.000000	0.166030	−0.034603	−0.203042	−0.766384	−0.204689	−0.207774
ECF	0.166030	1.000000	0.731569	0.555299	−0.506961	−0.535860	0.741805
ECE	−0.034603	0.731569	1.000000	0.403620	−0.501544	−0.255285	0.767138
ECW	−0.203042	0.555299	0.403620	1.000000	−0.033313	−0.006429	0.893724
ECW	−0.766384	−0.506961	−0.501544	−0.033313	1.000000	0.461798	−0.217048
ECB	−0.204689	−0.535860	−0.255285	−0.006429	0.461798	1.000000	−0.115588
Total	−0.207774	0.741805	0.767138	0.893724	−0.217048	−0.115588	1.000000

presenting the results. In the case of cumulative energy, correlation values r below 0.03 were calculated.

Additionally, those particular element (cumulative energy divided into 4 technological treatments) dependences are illustrated in Figure 3 (number of cows and cumulative energy in the scope of technological treatments), while the scope of the sources (elements) of cumulative energy intensity are shown in Figure 4.

The statistical analysis did not show any significant correlation between number of livestock units, and there was no correlation for any cumulative energy intensity values factors. In conclusion, it can be stated that the number of cows did not determine the level of the labor inputs or cumulative energy consumption in the scope of technological treatments or energy sources.

Among all the sources of cumulative energy tested here, the greatest part was the factors connected with machinery (fuel, mechanical and electrical energy intensity), resulting from the technological processes related to feeding and manure removal while using agricultural tractors. Milking, especially using automated milking systems (robots), required only a few percent of the total cumulative energy consumption.

4. Discussion

The capacity of the AMR automatic milking machine is approximately 90 cows/hour, and in this study, up to 75 cows were milked by 1 separate milking robot. This means that, theoretically, there is still potential for increasing the energy effectiveness.

A value of only 0.391 work-minutes∙LU⁻¹∙day⁻¹ for technical treatment I (milking and milk cooling) was observed for barn CF-AM with milking by four robots, AMS, which was the lowest inputs among all the tested barns. Similar results were presented by Freiberger, where labor inputs for milking at AMS ranged from 0.18 work-minutes∙LU⁻¹∙day⁻¹ to 1.06 work-minutes∙LU⁻¹∙day⁻¹, depending on the herd size (72 to 100 milk cows) [31].

According to Pommer et al. [32], with an average of 0.42 kWh per milking, the energy consumption of AMS is significantly higher than for conventional systems (0.28 kWh on average). In the barns examined in our experiment, the energy inputs for milking ranged from 0.648 kWh∙day⁻¹∙LU⁻¹ to 1.232 kWh∙LU⁻¹∙day⁻¹, which per milking gives from 0.297 kWh to 0.535 kWh. According to our findings, the data obtained from the experiments demonstrate the great advantages of automated milking systems, so the most sustainable solution could be indicated as automated milking together with a completely automated feeding procedure.

In our research, total energy inputs for all the technological treatments amounted to between 0.974 kWh∙LU⁻¹∙day⁻¹ and 3.309 kWh∙LU⁻¹∙day⁻¹. The use of a feeding robot almost completely eliminates the labor input for feeding concentrate feed, which may reduce its energy consumption. The more advanced the nutrition system is, the higher the energy inputs, as found in the study by Leola [33]. The results provided by our study differ from those presented in the literature [33]. In the barns with robotized feeding in our work, the human labor inputs amounted to only 0.838 and 2.020 for AF-AM5 and AF-AM7, respectively. Taking into account all the technological treatments with all the energy-related sources (machinery, equipment, electrical and mechanical energy, fuel consumption, labor inputs, and buildings), the most convenient solution was observed in the AF-AM7 barn with its full robotized systems of milking (four milking robots) and a TMR robot for feed preparation and distribution.

The cumulative energy intensity in the barns ranged from 19.344 for the AF-AM5 technology to 62.763 MJ∙day⁻¹∙LU⁻¹ for the barn with the lowest level of mechanization, where both milking and feeding was performed using conventional technologies, requiring heavy machinery, especially tractors, high fuel consumption and human labor services.

No comparison with other authors is possible in this matter. There is a lack of literature involving all the technological treatments in milk production, especially for the direct technological processes in barns, which were taken into account as a total, and this fact emphasizes the innovation of the research work conducted here.

The final results could help to indicate the best recommended technical solutions in sustainable milk production for owners of family farms. Further research could be performed on LCA analysis, taking into account the influence on the environment, level of advancement of the technology used, and social issues related to sustainable milk production.

5. Conclusions

1. Energy inputs for the production procedures amounted to 0.974 kWh∙LU⁻¹∙day⁻¹ and 3.309 kWh∙LU⁻¹∙day⁻¹. The majority of them were expenditures on preparing and serving feed. Labor expenditure ranged from 0.780 labor-minutes∙day⁻¹∙DJP⁻¹ to 3.410 labor-minutes∙day⁻¹∙DJP⁻¹, with the lowest related to treatment T_III—manure removal and storage.

2. Energy consumption accumulated in production activities, taking into account the energy consumption of the barn, amounted 19.344 to 62.763 MJ∙day⁻¹∙LU⁻¹. The lowest cumulative energy inputs occurred in the robotized barn in terms of milking and preparation and distribution of feed, while the highest energy consumption occurred in the barn using conventional methods of milking and feeding.

3. Automation and digital solutions in milk production involve increased energy inputs compared to traditional methods of mechanization. The solution to increased energy inputs in the case of milking robots or automatic feeding is the integration of the entire energy consumption system with its production system, e.g., through methane fermentation and then the production of energy from biogas.

4. Modern tools for forecasting energy consumption in animal production can be helpful in analyzing the energy consumption and creating model technological solutions that take into account modern solutions for automation and robotization of the production procedures.