Review of Seed Hemp (Cannabis sativa L.) Harvesting Techniques and the Challenges of Harvesting Technologies for This Crop

Abstract

Industrial hemp (Cannabis sativa L.) harvesting for grain represents a critical technological bottleneck in the modern supply chain, driven by a fundamental conflict between the plant’s resilient morphology and standard agricultural machinery. This review provides an analytical synthesis of harvesting methodologies, evaluating their performance against specific biological constraints such as extreme plant height (up to 4.5 m), high tensile fiber strength, and indeterminate ripening. Data synthesis reveals that hemp cutting is approximately 80 times more energy-intensive than for traditional forage crops, requiring an average maximum force of 243 N per stem. Comparative analysis demonstrates that while conventional whole-plant harvesting faces seed losses ranging from 26% to 46%, selective systems like specialized panicle mowers reduce these losses to nearly 2 kg·ha⁻¹ by targeting only the mature inflorescences. To ensure seed integrity and operational stability, the review identifies concrete technological priorities: the use of abrasion-resistant alloys for cutting edges, the implementation of non-binding shaft shielding (e.g., ABS piping), and a 40–50% reduction in threshing cylinder speeds compared to cereal settings. Future advancements must focus on specialized, high-clearance selective machinery and adaptive control systems to reconcile hemp’s unique physiology with industrial-scale efficiency.

1. Introduction

The high adaptability of industrial hemp (Cannabis sativa L.), stemming from its extensive cultivar diversity, facilitates its widespread cultivation across numerous global regions. Characterized by minimal agronomic input requirements and significant resilience to both biotic and abiotic stressors, these plants represent a compelling alternative for agricultural systems at various levels of advancement. Particularly in challenging environments where conventional crops underperform, hemp ensures consistent yields, contingent upon the strategic selection of cultivars tailored to specific production objectives [1]. However, despite its significant potential, the widespread industrial utilization of hemp is currently hindered by a critical technological bottleneck, primarily associated with the harvesting and primary processing stages. Harvesting hemp for grain involves distinct biological and engineering challenges that fundamentally set it apart from traditional cereal production. The absence of specialized, high-performance machinery remains the primary barrier to the efficient large-scale adoption of hemp in modern agriculture.

Modern hemp cultivation trends are strictly governed by the target quality parameters of the raw material. The production of fine fiber, intended for the textile industry, necessitates a precise agronomic approach, including early harvesting during the male flowering phase; this enables the acquisition of fibers characterized by high elasticity and a low degree of lignification. Conversely, harvesting focused on technical fiber—utilized in the biocomposites and natural building sectors—allows for later mechanical intervention, which results in higher biomass yields, albeit with increased rigidity and mechanical strength. This increased mechanical strength significantly raises the power requirement for cutting operations and accelerates the abrasive wear of machine components [2].

In recent years, there has been a growing interest in cultivating hemp for grain, driven by its increasing recognition as a valuable source of edible oil and protein-rich flour [3,4]. However, seed harvesting methodologies differ substantially from those employed for fiber-oriented crops. Harvesting hemp for grain involves distinct biological and engineering challenges that set it apart from traditional cereal production. Typically, the process entails either the selective harvesting of seed-bearing inflorescences (panicles) or the cutting of whole plants followed by threshing. Each approach necessitates specific technological frameworks and dedicated technical solutions, leading to the development of innovative machinery designs. These engineering responses are primarily driven by the plant’s unique morphology. For instance, extreme plant height—which can reach up to 4.5 m—necessitates the use of specialized, high-clearance headers to avoid the ingestion and processing of excessive amounts of woody biomass. Concurrently, the high tensile strength of the bast fibers frequently leads to the wrapping on rotating shafts and components, representing a major cause of mechanical blockages and significant operational downtime. Optimizing the performance and efficiency of these machines requires a compromise between environmental and biological factors—including morphological structure, ripening uniformity, and seed yield—and technical parameters aimed at achieving high levels of seed and stalk purity, minimizing crop residues, and preventing seed shattering, as the inherent asynchrony of seed ripening is a major mechanical constraint, forcing a difficult trade-off between harvesting immature seeds with high moisture content and losing mature grain due to shedding and cracking induced by the machine’s working units [5,6].

Simultaneously, the rising demand for renewable energy sources establishes hemp as a highly efficient energy crop. Hemp biomass, characterized by a high calorific value and a favorable combustion profile, is utilized in the production of pellets and briquettes, and serves as a substrate for biogas plants [7,8,9,10]. This provides the opportunity to manage both waste fractions from fiber production and whole plants cultivated in energy monocultures, both of which necessitate appropriate harvesting methodologies.

It should be noted that these are not the sole directions for hemp cultivation; a wide array of other potential applications is illustrated in Figure 1.

Contemporary applications of various industrial hemp components drive the demand for more efficient, mechanized, and high-performance harvesting systems. The diversification of end-uses necessitates fundamentally different harvesting strategies to optimize the quality of each specific raw material fraction. For instance, producing high-quality textile fibers requires early mowing and field-retting of the entire stalk to maintain fiber elasticity, whereas seed-oriented production focuses on grain maturity and necessitates selective systems to avoid the harmful inclusion of moist, resin-rich leaves and tough fiber. These systems must ensure the physical integrity of the harvested seeds while concurrently minimizing harvest losses in the field. Furthermore, dual-purpose cultivation, aiming for both seeds and fiber, relies on complex double-cut technologies that separate these fractions in a single pass to prevent cross-contamination and mechanical failures [6,11].

Despite this vast potential, the efficient utilization of hemp is hindered by technological barriers, primarily associated with the harvesting and primary processing stages. The specific physical attributes of hemp biomass, such as the high mechanical strength of the fibers and the presence of adhesive resins, impose unique demands on machinery systems. The specific physical attributes of hemp biomass, such as the high mechanical strength of the fibers and the presence of resins, impose unique demands on machinery systems. Consequently, the objective of this review is to provide a systematic classification of existing and emerging hemp harvesting technologies, with a specific focus on seed production. This study evaluates their operational performance and impact on seed quality and integrity in the context of the plant’s unique biological constraints. Finally, the paper aims to identify critical research gaps and outline technological priorities essential for optimizing harvesting processes and enhancing the economic viability of hemp cultivation.

2. Biological and Agronomic Constraints Influencing Hemp Harvesting Technologies

2.1. General Agronomic Considerations

Understanding industrial hemp harvesting technology necessitates, first and foremost, an analysis of the plant as a heterogeneous raw material. In contrast to cereal monocultures, where breeding and agronomy have long aimed for uniform ripening of the entire crop [12,13], hemp exhibits several ‘wild’ characteristics [14,15] that represent significant barriers from an agricultural engineering perspective.

A primary agronomic challenge arises from the frequent cultivation of industrial hemp for dual-purpose or multi-purpose objectives [1]. This necessitates harvesting technologies that can reconcile conflicting biological requirements: while seed collection requires advanced physiological maturity, the production of high-quality textile fiber is optimally achieved immediately following the pollen-shedding phase of male plants, prior to extensive lignification (woodiness) of the stalk. From a technological perspective, this incompatibility creates a significant engineering trade-off regarding harvesting time and machine configuration. Harvesting at seed maturity means that the fibers are already highly lignified, greatly increasing their strength. This results in higher energy consumption for cutting and a greater risk of fibers wrapping around the machine’s shafts compared to early-stage fiber harvesting. Furthermore, conventional threshing of whole, seed-ripe plants often leads to mechanical damage to the fibers, making them unsuitable for textile applications and limiting their use in technical products such as biocomposites or insulation. To address these conflicting requirements, dual-purpose machines must utilize independent functional modules, such as a “double-cutting” system. These systems utilize a specialized upper head for panicle processing and a separate lower unit that cuts woody stems for retting in the field, thereby minimizing the loss of fiber quality caused by delayed harvesting [16]. The machinery detailed in subsequent sections of this work represents a direct engineering response to this biological dualism.

Another crucial determinant is the growth dynamics and plant height. Cultivars grown in Central Europe can reach heights ranging from 1.5 to as much as 4.5 m (Figure 2) [14,17]. Furthermore, at the high sowing densities required for fiber production (up to 200–300 plants per m² [17]), such substantial biomass with high moisture content generates extreme resistance within cutting and conveying systems. Traditional cereal combine harvesters, typically designed for crops up to approximately 1 m in height, are prone to rapid clogging when processing such a stand structure. This limitation has necessitated the development of high-clearance machinery (e.g., the MultiCombine HC 3400, HANF FARM GmbH, Malz, Germany) and specialized headers that operate exclusively on the upper sections of the plants.

2.2. Stem Morphology

The morphological structure of the hemp stalk represents one of the most formidable challenges for harvesting machinery engineering, extending far beyond the scope of conventional solutions utilized in cereal or oilseed harvesting. Unlike the hollow and brittle stalks of cereal straw, the hemp stem possesses a unique architecture consisting of an outer layer of exceptionally strong bast fibers and a highly lignified (woody) inner core.

From a mechanical perspective, this composite structure is characterized by significantly higher cutting resistance compared to other agricultural crops. Research conducted by Chen et al. (2004) [19] demonstrated that the average maximum force required to cut a single hemp stem is approximately 243 N, with an average total cutting energy of 2.1 J. For comparison, the energy required to sever a hemp stalk is approximately 80 times higher than that required for typical forage crops. Furthermore, the energy demand is highly sensitive to the biological state of the plant; at a high moisture content (65%), the maximum cutting force increases by 51%, while the total cutting energy rises by 100% compared to stems at 8% moisture. Additionally, mechanical conditioning (maceration) of hemp stalks for field-retting is equally energy-intensive, requiring 10–60% more specific energy than the conditioning of alfalfa [19,20]. When harvested as whole plants, the stalks tend to entangle and wrap around machine components such as cutter bars, conveyors, and sieves. Such behavior accelerates the wear of these elements and leads to frequent blockages in combines and other harvesting equipment not specifically dedicated to hemp.

Furthermore, fibrous impurities impede the preliminary cleaning process in winnowing machines and stationary screens, as they have a propensity to form tangled aggregates that clog the perforations of the cleaning equipment.

To mitigate these operational issues, contemporary engineering responses emphasize a combination of material, structural, and mechanical adaptations [21,22,23]. Specifically, due to the high cutting resistance and the abrasive nature of the lignified hemp core, cutting edges and blades are increasingly manufactured from high-durability, abrasion-resistant alloys to significantly extend the operational life of the machinery. To prevent the “wrapping effect” caused by resilient bast fibers, rotating shafts are protected using ABS piping or smooth, non-binding sleeves, while exposed drive components are typically covered with high-density polyethylene (HDPE) panels or sheet metal deflectors to eliminate points of fiber accumulation. Furthermore, specialized hemp harvesters often feature extended hydraulic cylinders to increase header lift, which allows for higher cutting points and reduces the volume of woody biomass ingested into the machine. This is frequently complemented by the implementation of belt conveyors (draper headers) instead of traditional augers, a modification that significantly reduces the risk of fiber entanglement and ensures a more uniform and stable crop flow [19,23].

Consequently, engineering efforts are focused on designing systems—such as the aforementioned B-800 mower designed by Institute of Natural Fibers and Medicinal Plants—National Research Institute (INFMP-NRI), Poznań & FUGOR Sp. z o.o., Krotoszyn, Poland or dedicated stripping headers—that minimize the volume of stalk material ingested by the machinery. This approach ensures the preservation of the high physicochemical quality of the yield while simultaneously reducing the operating and maintenance costs of the machinery fleet.

2.3. Panicle Morphology and Seed Maturation

The selection of the optimal technology for harvesting hemp seeds and grain is determined by the plant’s unique botanical traits, which significantly deviate from the parameters typical of cereals or oilseeds. A primary challenge is the asynchronous ripening model (indeterminate ripening). Unlike cereals, which are characterized by a relatively short and uniform phase of technological maturity across the entire stand [12], the maturation process of hemp seeds within the panicles progresses successively along the inflorescence axis. While seeds in the lower whorls of the panicle reach full biological maturity—characterized by a hard seed coat and maximum accumulation of reserve materials—the apex of the inflorescence may still be at the milk stage. This physiological disparity necessitates the determination of a compromise harvest date, typically occurring when 70–80% of the seeds in the stand exhibit signs of maturity. An error in assessing this timing generates twofold losses: premature harvesting results in a high proportion of immature seeds with a low thousand-seed weight (TSW), an underdeveloped fatty acid profile, and reduced germination energy. Conversely, delaying the harvest drastically increases the risk of losses due to the natural shattering of the most mature seeds [11,24]. This inherent biological asynchrony directly dictates the suitability and efficiency of various harvesting systems. Conventional whole-plant harvesting followed by windrowing is particularly disadvantaged by this trait, as the multiple handling steps required for the drying crop significantly exacerbate shattering losses of the most mature, loosely attached seeds. In contrast, single-pass combine harvesting of low-growing varieties (e.g., ‘Henola’, ‘Finola’) attempts to mitigate shattering by immediate threshing, yet it often faces challenges associated with the high moisture content of immature seeds and green leaf fragments, which can quickly compromise the quality of the entire lot if not dried within 2–4 h. Selective harvesting systems, such as panicle mowers and stripper headers, represent the most effective engineering response to indeterminate ripening. By targeting only the upper, more mature sections of the stand and utilizing sealed collection tanks, these systems not only capture shattered seeds that would otherwise be lost to the ground but also minimize the intake of moist, green biomass from the lower parts of the plant, thereby enhancing both seed homogeneity and storage stability. These losses are further compounded by any mechanical interaction between the machine’s working components (e.g., the reel or cutter bar) and the plant before the material is introduced into the threshing system [25].

The mechanical harvesting of industrial hemp is further complicated by the specific morphological structure of the panicles. The seeds, embedded within dense inflorescences, are surrounded by bracts (perigonal leaves) that often remain physiologically active and exhibit high moisture content at the stage of technological maturity. The presence of these bracts, along with fine fiber fractions, woody core particles, and leaf fragments in the threshed material, directly influences yield parameters immediately after cutting. These components possess significantly higher hygroscopicity than the seeds themselves. Due to their porous structure, hemp seeds rapidly absorb this moisture, leading to increased hydration even under favorable weather conditions during harvest. This phenomenon drastically reduces the safe storage window for the threshed mass. To prevent the degradation of material quality, stationary drying must commence immediately: within 3–4 h post-harvest for industrial grain, and within a maximum of 1–2 h for sowing material [5,11,26,27].

Hemp seeds exhibit extreme sensitivity to dynamic loading. The grain is relatively susceptible to fracturing under the impact forces encountered within the threshing drum. To preserve seed integrity, machine adjustment parameters must be strictly optimized to mitigate these impact forces. Specifically, this necessitates a significant reduction in threshing cylinder speed—often to levels 40–50% lower than those used for small grains—to minimize the kinetic energy of the rasp bars during contact with the seeds. Furthermore, the concave clearance must be widened to accommodate the bulky, resinous panicle material, preventing the seeds from being crushed against the concave wires during the threshing process. Additionally, since hemp seeds are relatively light and possess a different aerodynamic profile than cereals, the cleaning fan speed must be precisely reduced to prevent the seeds from being discharged with the chaff [22,23]. Even microscopic damage to the seed coat (testa) occurring during intensive threshing—for instance, as a result of excessive cylinder speeds—carries grave implications. The resulting oil leakage triggers rapid oxidation and rancidification, which renders the raw material unsuitable for food-grade applications. Furthermore, such structural damage provides a point of entry for fungal pathogens, facilitating mycotoxin contamination during storage. Ultimately, the combined impact of mechanical injury and pathogen pressure leads to a significant decline in the emergence rate and germination vigor of the sowing material [5].

3. Review of Industrial Hemp Harvesting Methods

Globally, approximately two thousand plant species are known to yield natural fibers, yet only a few are of commercial significance, accounting for nearly 90% of global natural fiber production [28]. One of the most prominent among these is hemp, which, in addition to the aforementioned fiber, yields seeds as a critical product. The production of high-quality seed, seed-derived oil or fiber depends primarily on the quality of the harvested raw material. This, in turn, is related, among other factors, to the harvesting technology used and its efficiency. In most instances, these technologies are developed locally, building upon existing technical solutions tailored to specific regional agricultural practices [11]. Some of these technologies can be used on small areas of land due to their efficiency in relation to the size (working width) of the harvesting machine that cuts the plants.

The agrotechnical timing of harvesting is important for the subsequent desired use of hemp seeds. To ensure the efficient extraction of essential oils, the inflorescences (or floral biomass) should be harvested during the full flowering stage, when the concentration of secondary metabolites is at its peak. Conversely, when the crop is designated for grain or sowing seed production, harvesting must be conducted at the physiological maturity stage, with the precise timing contingent upon the specific harvesting methodology employed [11].

Hemp harvesting technologies can be implemented through various methods, depending on the intended end-use of the harvested products. This diversity in harvesting approaches necessitates the application of various machinery and equipment, which, despite their different functions, often share similar structural and design characteristics. Harvesting hemp specifically for seed production can be categorized into the following methods [5,11,28,29,30]:

• Whole-plant harvesting
• Segmented (cut-to-length) harvesting
• Selective harvesting.

The above-mentioned hemp seed harvesting technologies are characterized in

Table 1. Comparison of the methods used at each hemp harvesting technologies.

Technology	Methods	Shearing Units	Preferred Crop Size *
Harvesting of whole plants	Cutting plants and arranging them in swaths	Sickle-bar mowers, cutter bars	Small- and medium-sized
	Cutting plants and collecting them in the machine’s hopper	Sickle-bar mowers	Small sized
	Cutting with simultaneous seed threshing (harvesting with adapted combine harvesters)	Sickle-bar mowers,	Large- and medium sized
Harvesting of plants divided into sections	Cutting entire plants and arranging stem sections in swaths	Sickle-bar units operating at two or three height levels (fixed or adjustable)	All sizes
	Cutting entire plants and arranging stem sections in swaths	Drum cutting units	Large- and medium- sized
	Double-cut system: collecting panicles (and potentially leaves) while arranging stem sections in swaths	Reciprocating units (upper) and rotary units (lower)	Large-sized
		Combing, stripping (upper unit) and rotary cutting assemblies (lower unit)	Large-sized
Selective harvesting	Cutting off panicles, inflorescences and leaves; leaving uncut stems in the field for subsequent harvest	Sickle-bar units, combing and stripping units	Large- and medium- sized

Source: Own elaboration. *—Small denotes cultivation areas up to 2 ha, medium denotes 2–5 ha, and large denotes exceeding 5 ha.

.

Harvesting whole plants is nothing more than cutting entire stems in a single pass. In this context, two technological approaches can be employed. The first is single-stage harvesting using specially adapted combine harvesters. Notably, leaving threshed seeds in the machine’s grain tank for extended periods may cause them to heat up significantly, reducing their germination capacity and increasing the risk of fungal infection. This results in a significant loss of the seeds’ valuable biological and chemical properties. The second approach is a two-stage harvest, which involves cutting the plants in the first stage and leaving them to dry in the field, followed by threshing using stationary seed separators or combine harvesters.

Another group of technologies involves harvesting hemp and cutting it into sections. The machines used in this technology performs the following operations:

• cutting whole hemp plants;
• cutting them into sections approximately 1 m in length, with the option of separating the panicles;
• collecting and transporting the cut plants for on-farm drying, or
• leaving the segmented plants in the field to dry;
• threshing the seeds and processing the stems.

The optimal technology for the harvesting and primary processing of seed hemp should be conducted according to the following sequence:

• Mowing and collection of the hemp panicles alone;
• Drying of the panicles using either heated air or natural ambient conditions;
• Threshing of the pre-dried panicles.

In this scenario, the mowing of the hemp stalks is performed as a subsequent operation. This approach is formally defined as selective harvesting.

In their research, Manea et al. [31] presented a comprehensive schematic overview of harvesting technologies for hemp stalks intended for fiber production. The classification clearly categorizes harvesting methods into two technological streams, one of which focuses on seed harvesting. However, the authors do not detail this specific method extensively. The second technological direction identified involves whole-stem harvesting. Within this category, they describe various machinery configurations and technological possibilities for process implementation, some of which are discussed in detail later in this article. Furthermore, the authors distinguish a specific technology based on a machine developed by their team for mowing and windrowing hemp plants for drying its construction and operational analysis are presented in their subsequent work.

Additionally, Gusovius et al. [28] developed a conceptual framework illustrating the process sequences and technologies identified through their analysis of established industrial hemp harvesting procedures. The authors indicate that hemp can be harvested for seeds either through direct threshing during cutting or by threshing post-drying in swaths. Subsequent stages of the proposed technologies include pre-drying hemp straw in order to facilitate transportation to processing facilities (e.g., for fiber extraction). They also highlight the possibility of transporting cut plants to drying facilities for further processing, depending on the specific plant fraction. Similar classifications of harvesting methods and technologies have been proposed by other prominent researchers and institutions, including [11,21,22,23,32,33,34].

3.1. Whole-Plant Harvesting

Whole-stalk hemp harvesting technologies are primarily governed by cost-minimization principles. Mown stalks—either intact or stripped of leaves, inflorescences, and seeds during the cutting process—are typically left on the field surface, either in their full length or in long sections, for field curing and subsequent collection for further processing. The material obtained via this method consists of long fibers, which are highly valued for textile applications. This system is not extensively adopted, and specialized machinery for this type of harvest remains limited [11,21,22,31,35,36,37,38]. Nedelcu et al. [38] proposed an integrated harvesting methodology encompassing the development of equipment designed to cut the stalks, bind them into sheaves, and deposit them on the ground.

The only known commercially produced machine developed for whole-stalk harvesting is the ZK-1.9 (SEL’MASH, Bezhetsk, Russia). It was specifically engineered to cut and align the stalks parallel to one another in orderly windrows. The machine could be equipped with an attachment that converted it into a type of sheaf-binder, enabling the formation of bundles that were deposited on the field surface. These bundles were subsequently manually stacked for air-drying [11,21,25]. This method of drying hemp stalks is recommended, for instance, by the United States Department of Agriculture [39].

Finger-type or double-knife cutter sickle-bar windrowers have also been employed for the harvesting of whole hemp plants [40]. This technology offers distinct advantages, specifically by minimizing seed shattering losses typically induced by wind or the mechanical impact of the reel. The mown crop underwent a natural field-curing process prior to collection, which effectively lowers the energy requirements and technical demands of subsequent post-harvest drying [23].

While this method of harvesting is applicable to large-scale cultivation, its success depends on level terrain, adequate solar exposure and favorable wind conditions to facilitate drying. Furthermore, the suitability of the combine harvester for threshing the dried material must be ensured. This necessitates additional adaptations and associated costs, such as the implementation of a pick-up header, increased concave clearance, and reduced rotational speeds of the functional units to prevent the entanglement of long hemp stalks.

Mowing can be carried out at a seed moisture content of 15–18% before the ripe seeds fall out of the husks. Mowing at a seed moisture content below 15% leads to increased losses due to shattering. Under favorable environmental conditions, seed moisture can drop to 10–12%. Drying whole plants in swaths can also be considered a risk management tool against losses caused by strong winds. It should only be considered when the weather forecast predicts dry weather for the next two to five days, so that harvesting can be completed before the rain arrives.

The advantages of swath drying are:

• Reduction in wind-induced losses.
• Partial drying of seeds before harvesting.
• Potential reduction in aeration and drying requirements.

It is possible to increase drying efficiency by tedding the swaths and then raking them. These technological processes generate certain costs related to both their implementation and the appropriate preparation of machines for working with long-stemmed plants; for example, addressing exposed rotating structural elements on which fibers can wrap around, even those that do not directly participate in the technological processes. During these operations, there is also uncontrolled seed shattering when the plants are turned, combined with the panicles hitting the machine components. In this way, more than 20% of the seeds may be lost, falling onto the surface of the field [23].

Its main disadvantage is primarily the difficulty of mechanizing the entire process from mowing to harvesting seeds, leaves and fiber. In addition, the mowed plants lie for several days directly on the surface of the field, during which time they are in constant contact with the ground and soil. Depending on the species of hemp, its stems can range from several dozen centimeters to over 2 m in length and have a diameter of up to several millimeters at the base, which means that with a high yield, the mown swath can be up to several centimeters high. In such a case, drying the plants from the lower levels of the swath will be very difficult. Restricted airflow, and thus the lack of moisture evaporation from the plants, combined with the constant contact of the stems and seeds with the ground, will expose them to the development of mold and bacteria and contamination with coliform bacteria. Such effects may result in a loss of quality and a reduction in the grade of seeds and stems, and consequently in the quality of products obtained from them, such as oils or fiber, which entails obvious economic losses for plant producers and processors.

The primary disadvantage of this method is the significant difficulty in fully mechanizing the entire production chain, from initial mowing to the eventual extraction of seeds, leaves, and fiber. Furthermore, the direct contact of the seeds with the soil surface exposes them to the risk of fungal growth and bacterial proliferation. This poses a particular threat of contamination with coliform bacteria (e.g., Escherichia coli), ultimately leading to quality degradation and the downgrading of the seed grade [23].

A renewed attempt to implement whole-plant harvesting—while avoiding in-field windrowing—was undertaken at Nanjing University in China. A prototype harvester was developed (Figure 3) [35], which features a finger-type cutter bar mounted at the base of the frame and several chain conveyors equipped with lugs (grippers). Transverse conveyors, integrated with row dividers, initially support the plants to facilitate precision cutting and subsequent transport in a vertical orientation. A longitudinal conveyor then transfers the harvested plants into a hopper. Due to its status as a prototype, the machine’s field capacity is relatively limited, ranging from 0.15 to 0.22 ha·h⁻¹ [35].

While whole-plant harvesting undoubtedly facilitates the production of high-quality seeds, this advantage is offset by the necessity for additional handling and significantly higher operational costs. Plants harvested via this method must subsequently be dried, followed by panicle separation and threshing. Several comparable machines have been described in the research by Kirbaş et al. [34].

3.2. Segmented Harvesting (Cut-to-Length)

The objective of mechanizing the harvesting process for industrial hemp, combined with the pursuit of maximizing the yield and quality of seeds and leaves, has driven a transition from whole-plant harvesting to methods involving plant segmentation. This technology evolved from harvesting of whole plants. One of the main reasons for its development was the existing technical and technological capabilities for producing fiber from stems. Hemp is the tallest cultivated fiber plant, as some cultivars can reach up to 3 m in height while others barely exceed 1 m. Such diversity in plant length makes it difficult to standardize the technology for extracting fiber from their stems. This technological trajectory is significantly influenced by the requirements of the downstream processing industry. The capacity to process full-length stalks is restricted, as existing industrial processing lines are generally designed for various fiber crops, such as flax or kenaf, which typically reach heights of slightly over 1 m. Consequently, hemp processing in Europe frequently utilizes technological lines originally dedicated to flax, which are configured to handle stalks with lengths generally not exceeding 1 m [41].

Cut and segmented stems can dry out on the surface of the field alongside the cut and swathed panicles. More frequently, however, the panicles are collected for transport to drying facilities, which significantly improves the quality of the harvested seeds and reduces losses associated with shattering. This harvesting method facilitates the easy and cost-effective separation of the seed-bearing panicles for independent processing (i.e., threshing). One of the initial approaches to harvesting whole seed-hemp plants, specifically oriented toward the high-efficiency recovery of high-quality seeds, was a machine developed at theINFMP-NRI in Poznań. Technological research led to the design of a tractor-drawn hemp harvester designated as the “KR” model, which is capable of the simultaneous mowing of entire plants and the detachment of the panicles [11]. The machine performs three concurrent operations: mowing the stalks, segmenting them into three sections, and severing the panicles from the stalks. The detached panicles are then transported via a belt conveyor to an integrated agricultural trailer. Meanwhile, the segmented stalks are left on the field for curing; once partially dried, they can be collected for further fiber processing using standard equipment, such as round balers. Figure 4 illustrates the design of the machine. Given its operating parameters, the “KR” harvester is specifically adapted for small- to medium-scale cultivation (up to several hectares), reflecting the typical plantation sizes found in Poland and South-Eastern Europe [11]. The machine had a working width of 1.9 m and could operate at a capacity of 2.4 ha·h⁻¹.

Another technologically straightforward hemp harvesting system involves segmenting the stalks into lengths of approximately 70 cm through the use of machines equipped with multiple (typically 2–3) cutter bars operating at staggered heights above the field surface. Their spatial arrangement is engineered to produce stalk fragments that meet the specific technical requirements of processing facilities. The cutter bars are mounted on the machine frame in a configuration that ensures sequential cutting from the top of the plant downward. This tiered arrangement facilitates a uniform distribution of the severed stems and, crucially, enables the collision-free detachment of successive vertical sections of the crop. Consequently, the harvested inflorescences and panicles fall to the ground, where they mix with the stalk segments. In this state, the biomass is left for field-curing and subsequent collection for further processing.

An example of this technology is the harvester shown in Figure 5, which is referred to in the literature under several designations: Clipper 4.3 MMH (Tebeco, Prague, Czech Republic) [11] and Tebeco Beagle 3.3 [40]. The Tebeco company also introduced a model equipped with two cutter bars, the Tebeco Beagle 3.2 [40]. The former machine can operate at an operational field capacity of 4–5 ha·h⁻¹, while the latter achieves a capacity 4 ha·h⁻¹ [11,40].

According to research conducted by Ivanovs, at that time (2008), the utilization of this machinery was considered economically viable for cultivation areas exceeding 50 ha [40].

A similar operating principle is employed by a machine developed in Romania [31], as illustrated in Figure 6. This device was specifically engineered for use on small- and medium-sized farms [37]. The operational process involves the simultaneous cutting of the inflorescences via an upper cutter bar with a hydraulically adjustable working height, while the remaining “whole” green stalks are severed by a lower cutter bar. Both units utilize finger-type cutter bars featuring a scissor-action cutting system. A comparison of the primary technical parameters of selected harvesters is summarized in

Table 2. Comparison of technical characteristics and performance parameters of three different hemp harvesters.

Cutter	Maximum Working Width	Working Speed	Area Performance	Swath Width	Power Requirement
	mm	km·h−1	ha·h−1	mm	kW
KR mover	1900	11–12	2.4	700–800	22.6–35
Tebeco Beagle 3.2	3000	12–13	3.9	3000	35–50
INMA mover	1300	5.5–6.5	0.9–1.0	1300	45–50

Source: Own elaboration based on [11,31,40].

.

The principle of tiered, multi-level cutting for tall seed hemp plants was also utilized by Bavarian engineers in Germany, who developed machine prototypes designated as HMG 4-24, SMU-2, and SMU-3-210. Concise, synthetic descriptions of these designs—excluding technical specifications—were included in a report by the IMU-Institut Berlin. Another German manufacturer of special-purpose machinery, Kranemann GmbH (Klocksin/Blücherhof), also developed and tested a hemp harvester designed to deposit mown stalks on the field in a cross-parallel arrangement [28]. Similarly, the Lithuanian company Laumertis produces dual- and triple-tier machines for tiered hemp harvesting. These machines operate with a working width of 2.9 m, capacity of 2 to 7 ha·h⁻¹ and require a tractor with a power output of 150 kW [42].

Another, earlier version of the technology for cutting hemp stalks into long sections involves the use of a drum assembly for cutting and dividing them. In this case, the blades were mounted on three levels of two co-rotating drums, each with a diameter of approximately 1.0 m. This arrangement allowed plants to be pulled into the gap between the drums, which was twice their diameter, and simultaneously cut and divided into two sections. The cut and segmented plants were then laid in swaths between the wheels of the self-propelled power unit. This was the functionality of a German machine called ‘Bluecher 02/03’ manufactured by Kranemann (Klocksin, Germany). Developed in the late 1990s, this machine was built upon the platform of a New Holland 1905 forage harvester (Figure 7) [11,43,44].

Used as the power unit for the aforementioned harvester, its 221 kW engine made it significantly more expensive and less efficient than operating a tractor with sickle-bar machines.

The primary rationale for utilizing this harvesting system was the objective of maintaining the original vertical orientation of the hemp plants until they are segmented into lengths of 600–700 mm. During the integrated mowing and cutting process, the conveyor components capture the stalks and stabilize them in an upright position. Subsequently, cutting disks mounted at fixed intervals on the drum partition the stalks into sections, while the plant material maintains its natural vertical orientation throughout the entire operation. Finally, the segmented stalk fragments are deposited on the field surface in an orderly windrow [11,21,28,34,43,45].

As already mentioned, when using this type of machinery, particular attention must be paid to plant height. Depending on the variety, seed hemp can reach heights ranging from 1 m to as much as 3 m [14,17,18,44]. For tall varieties, it is evident that a lack of adjustable cutting heights—with the exception of the machine developed by Romanian [31,46]—results in the production of stalk segments of inconsistent lengths. In harvesting systems based on cutter bars, the mown material is distributed across the entire field surface (wide windrows); this ensures that both field retting (maceration) and drying are uniform and accelerated [11,21,31,40]. In the case of a drum system, the swath is laid in the center of the machine’s working width, which makes it significantly thicker. In this case, field drying was difficult, and to achieve the desired effect, it was necessary to use straw walkers.

A significant drawback of this system arises when the stalks are no longer upright due to wind or lodging; in such instances, the length of the stalk segments increases, which leads to complications during downstream processing [45]. Furthermore, when inflorescences and stalks are dried in windrows, rainfall can cause quality degradation and trigger premature germination within the windrow. There is also a substantial risk of contaminating seeds and inflorescences with stones, soil, foreign matter, or manure. It is important to note that standing plants dry more rapidly in humid conditions than mowed crops lying in windrows [23]. In seed-oriented harvesting, there is also a considerable risk of increased seed loss due to natural shattering; this is particularly critical when hemp is harvested at the full maturity stage.

The rapid advancement of agricultural engineering and technology over the past two decades has significantly impacted hemp harvesting, particularly regarding high-efficiency yields and the recovery of high-quality seeds with minimal shattering losses. Conventional mowers, cutters, and similar equipment discussed previously are increasingly being replaced by combine harvesters or specialized machinery designed for the selective harvesting of inflorescences and seeds and stems for fiber. The latter will be described in greater detail in the subsequent section of this article.

Currently, several types of specialized hemp combine harvesters are in operation. During the harvesting process, these machines typically facilitate the separate collection of inflorescences and seeds, and in some cases, leaves and stalks separately [5,11,21,22,30,34,35,43,47].

In general terms, combine harvesting of hemp involves a single-pass operation where a single machine (the harvester) performs the mowing of the plants, which are subsequently segmented into longer sections, followed by the separation of the inflorescences, seeds, and/or leaves from the stalks. The separated inflorescences are either threshed or accumulated in tanks, which may be integrated into the harvester or situated on trailers traveling alongside. A similar process applies to the leaves, which are frequently transported together with the inflorescences. The stalks, after being cut and segmented, may be left in a disordered manner in a random orientation, forming windrows or uniform layers of mowed material. These windrows are subsequently collected using standard agricultural balers, such as round balers or large square balers (e.g., the Hesston type). However, it should be emphasized that fibers obtained from plants harvested in this manner may exhibit lower quality, which complicates their application in the textile industry [45]. This quality reduction is driven by the same factors previously identified in windrow-based harvesting systems. In certain technological approaches, entire plants—dried in the field prior to mowing—are shredded and transported by the harvester’s conveyor systems into trailers traveling alongside [22].

Combine harvesting of seed hemp has been conducted for over 20 years using the “Double-Cut system,” which was developed and implemented by the Dutch company HempFlax [11,21,28,34,43,48]. The harvest system’s name is derived from the integration of two separate headers into a single machine. The upper header is height-adjustable, controlled via a hydraulic lift. Depending on the harvesting objective—be it seed quality and yield, or more recently, leaf recovery—this unit is either a traditional grain combine header or a specialized stripper header designed to harvest only the inflorescences. In these technical configurations, the upper header typically delivers the harvested biomass to a conveyor, which transports it to the combine’s threshing unit, an integrated storage tank, or a trailer (either towed or side-loading). The second, lower header is typically a modified forage harvester header (e.g., Kemper rotary type). It cuts the lower stalks at a height of several to a dozen cm above the ground and feeds them into a central section equipped with drums for the pre-maceration (conditioning) of the cut stalks. This process facilitates faster field drying. Subsequently, the stalks are generally discharged beneath the machine and deposited in a windrow.

HempFlax, based in Oude Pekela, the Netherlands, was one of the pioneering companies to develop this type of machinery. Several years ago, GroeNoord and HempFlax joined forces to establish CANN BV, aimed at providing optimal solutions for the hemp harvesting sector. CANN, an acronym for Crop Ability Noord Nederland, has since established itself as a leading specialist in the niche market of hemp harvesting technology [48].

Their initial machinery, dating from the late 1990s and the early 2000s, operated in a single-header configuration. These systems were designed either as implements for high-horsepower agricultural tractors or as specialized attachments for self-propelled forage harvesters, such as the Claas Jaguar 800 series [43].

The machine design was refined by Wittrock company (Rhede, Germany) and introduced to global markets under the designations HempCut 3000 or HempCut 4500, depending on the working width. The HempCut features a modified header developed by Kemper (Stadtlohn, Germany) and a customized cutting drum equipped with a single knife. During operation, the plant stalks are fed longitudinally into the cutting drum, where they are segmented into sections of 600–700 mm and discharged directly beneath the drum into a windrow. The operational capacity of this machine ranges from 2.1 to 3.4 ha·h⁻¹ at an operating speed of 5–10 km·h⁻¹ [43]. It is therefore suitable for harvesting larger areas.

One of the most noteworthy iterations of this combine harvester system for harvesting hemp, specifically optimized for seed recovery, has been developed by HempFlax in the Netherlands. Known as the “double-cut system,” it integrates a conventional HempCut 4500 unit with a modified combine harvester header. The HempCut 4500 system has been adapted to various carrier vehicles. Perhaps the most compelling configuration involved the use of a chassis from another Dutch manufacturer, Agrifac, which was originally designed as a sugar beet harvester platform (Figure 8) [30,34].

This specific design, as well as several other machines utilizing the “double-cut system,” does not perform seed threshing on-board. Instead, these harvesters are equipped with a conveyor—located on either the right or left side, depending on the operator’s cab access point—that transports the harvested panicles and leaves directly into a storage container [49]. The lower row-crop header severs the stalks and deposits them in a windrow for field drying. In the case of the machine illustrated in Figure 8, the delivery of the cut panicles is facilitated through a discharge chute, analogous to the spout mechanisms employed in forage harvesters [30]. However, it would be advisable to standardize this design to facilitate more and, above all, comparable studies of this hemp harvesting technology.

Numerous other examples of such adaptations and modifications can be identified. Hemp Farm, a New Zealand-based hemp production and processing company, utilized a Claas Lexion 670 combine harvester as the platform for its double-cut technology. In this configuration, a Claas 600-series row-crop header was installed as the lower harvesting unit to sever the stalks remaining after panicle removal, windrowing them between the harvester’s wheels. The upper harvesting unit, a MacDon D120X draper header (notably featuring belt conveyors), feeds the material centrally. From there, the biomass is moved via the feeder house into a drum and flail threshing assembly. Threshing residues are deposited in a separate windrow at a short distance from the stalk windrow. To facilitate this, a belt conveyor was mounted behind the straw walkers to discharge the residues to the left [50]. Such a precise separation of fractions allows for the subsequent collection of stalks for fiber processing without unwanted contamination from seed threshing residues.

An early modification of this configuration introduced by HempFlax involved replacing the conventional auger conveyor in the upper header with a belt conveyor. This system transports the severed panicles to the side of the machine, depositing them into a single lateral windrow. Meanwhile, the stalks cut by the lower row-crop header—segmented into lengths of 60–70 cm—form a second windrow positioned between the harvester’s wheels. Consequently, the machine produces two distinct windrows: one comprising the lower stalk segments and the other consisting of the upper plant sections (Figure 9).

Since 2012, this system has been utilized for the harvesting of both seeds and leaves. Following a brief field-drying period, the upper stalk sections containing the seeds can be collected and threshed using a standard combine harvester equipped with a conventional grain header [11,21,23,28,34,43,48].

Following the establishment of CANN BV, a modern harvester based on the John Deere T6 series combine platform was developed. The newly designed dual-header system is available in several versions and combinations. In the base configuration, the upper unit consists of a standard John Deere 616R grain header or a specialized CANN 616 Draper stripper header. The lower unit, meanwhile, is a modified forage harvester header adapted for hemp—either the John Deere 445 or the Kemper 455 rotary header (Figure 10) [51,52].

In conjunction with the TFT10 leaf harvesting trailer, the DC4510 DoubleCut module enables high-quality separation of all three primary hemp plant components—seeds, biomass, and fiber—within a single-pass operation. A working width of 4.5 ensures an optimal windrow for the field-drying process. Under optimal conditions, this system achieves an operational capacity of 2.5 ha·h⁻¹. The upper header is adjustable, allowing for plant harvesting at heights ranging from 1.2 m to 3.2 m. Depending on the specific requirements, the machine can operate using either one or both headers.

The harvested tops are delivered via a specialized feeder to the threshing unit, while the biomass (leaves) from the straw walkers falls onto a secondary conveyor for transport to the integrated TFT10 high-capacity trailer. This trailer can accommodate 2–3 t of inflorescences and leaves. Simultaneously, the lower header cuts the stalks and feeds them centrally into counter-rotating drums, which pre-macerate the material before discharging it onto the field in a windrow for drying [51,52].

Furthermore, the company offers technology dedicated specifically to fiber harvesting: the CANN SF4510, built upon the John Deere 8400i forage harvester platform and integrated with a custom-adapted Kemper 455 header. Operating with working widths of 4.5 or 6.0 m, this machine provides optimal windrow dimensions for field drying, facilitating efficient and high-quality baling. The Kemper unit has undergone several modifications to prevent the wrapping and jamming of hemp stalks; the original cutting head frame was replaced with a custom frame and drum, allowing for an increased cut length of up to 50 cm. This custom configuration also enhances crop flow, resulting in high throughput while maintaining superior windrow quality for subsequent raking and baling [52]. Italian researchers have also investigated the use of the Claas Jaguar 890 SPFH for similar harvesting objectives. The machine moved at a speed of 4.0 km h⁻¹, which corresponded to a working capacity of 2.08 ha h⁻¹ [22].

Russian scientists and engineers are also conducting research on the development of specialized hemp combine harvesters, focusing on the simultaneous recovery of fiber and seeds. For fiber-oriented harvesting, they modified a Don-680 (Rostselmash) forage harvester by integrating a specially designed header with a functional principle similar to the Kemper 455 (Figure 11). For seed harvesting, they utilized a Don-1500B (Rostselmash) grain harvester, for which they also developed a proprietary dual-header system [47].

In the case of the Don-1500B combine, the upper header is configured as a stripper unit. The panicles recovered by this mechanism are delivered to the threshing assembly for processing. Simultaneously, the lower header cuts and pre-macerates the stalks before depositing them in a windrow. For the forage harvester attachment, operating with a working width of 4.5 m and at speeds ranging from 5.0 to 10.0 km·h⁻¹, an operational capacity of 3.5–5 ha·h⁻¹ was achieved. In contrast, the Don-1500B combine, utilizing a 7 m working width at operating speeds 2.0–3.0 km·h⁻¹, demonstrated a capacity of 1.0–1.5 ha·h⁻¹ [47].

In its report for growers, the Government of Alberta [23] recommends the use of rotary combine harvesters (e.g., New Holland models) for harvesting low-growing seed varieties (up to 1.25 m) such as Finola. This specific type of threshing system facilitates effective seed processing while minimizing the release of long fibers, which tend to wrap around internal components and disrupt the machine’s operation. For proper operation, the combine harvester should undergo certain modifications to protect its working parts—especially the moving and rotating ones—from becoming entangled with fibers.

○

Shielding: Exposed moving parts should be shielded with high-density polyethylene (puckboard) or sheet metal to prevent fiber wrapping.

○

Deflectors: Deflectors should be installed to keep the crop away from the header reel ends and to narrow the feeder house inlet, preventing fiber from wrapping around outer shafts and pulleys.

○

Shaft Protection: ABS piping can be placed over front drive shafts to provide a smooth, non-binding surface that prevents fiber entanglement.

○

Line Management: Cables and hydraulic lines should be secured tightly to the machine’s chassis to reduce the risk of fiber build-up.

○

Header Height: Header hydraulic cylinders should be extended by approximately 30 cm to allow for higher lift, which is essential when handling taller varieties.

Researchers from the Italian institute CREA (Consiglio per la Ricerca in agricoltura e l’analisi dell’Economia Agraria) have been involved for years in developing new technologies and solutions for emerging supply chains in the non-food sector [5,53,54,55]. As part of this research cycle, they attempted to utilize a modified New Holland CR 9080 combine harvester for hemp harvesting, aiming to obtain high-quality sowing seeds while minimizing seed losses.

These studies were conducted on the Futura hemp variety, sown in late June in northern Italy and harvested in October using the aforementioned combine, which was equipped with a specially developed separator for recovering threshing residues. The shortened growth cycle did not significantly affect plant height, which averaged 173 cm. The effective operating time of the combine accounted for 57% of the total working time, and the field capacity reached 1.14 ha h⁻¹ for the simultaneous harvesting of entire plants and seed threshing. It was determined that seed losses resulted primarily from the performance of the mowing and threshing systems but remained below 5%. Furthermore, the developed separation system enabled the recovery of 492.20 kg ha⁻¹ (on a dry matter basis) of high-value threshing residues [22].

Another Dutch company involved in hemp cultivation and processing, Dunagro, has also developed its own machine operating on the “double-cut” principle. This harvester, named the Hempbull (Figure 12), is built on a Claas Xerion system tractor chassis. In this configuration, the plant tops are harvested by an upper stripper header (Figure 13) and transported via a belt conveyor into an integrated loading bunker. Simultaneously, the lower Claas row-crop header cuts the stalks remaining after the removal of the panicles and leaves, depositing them in a windrow between the machine’s wheels [56].

In this approach, the upper header severs the plant tops, which are then either delivered directly to a storage tank or fed into the combine’s threshing unit, provided the machine is based on a grain harvester platform. The second, lower header cuts the remaining stalks and deposits them into a windrow for field drying. This method, when integrated with on-board seed threshing, is highly efficient and minimizes the number of required field operations. However, it necessitates extensive internal modifications to the combine to prevent seed damage and minimize losses. These modifications typically include the installation of gentler threshing drums, wider concaves, and modified sieves, as well as significantly reduced fan speeds. The best results are achieved when the crop exhibits uniform maturity and the plants are of moderate height [22]. Despite these technical requirements, this harvesting method facilitates the recovery of high-quality hemp seeds with substantially lower losses compared to traditional methods that involve cutting and leaving the entire plant in the field [11,22,28].

The advantage of this system lies in its ability to effectively separate stalks from seed panicles. This segregation high-quality seeds and facilitates superior field-drying of the stalks. Because the resulting swaths consist exclusively of stalks and are thinner, moisture dissipation into the atmosphere is significantly accelerated. Consequently, this technology enables the recovery of high-quality hemp stalks suitable for downstream processing into various fiber-based products. This double-mowing technology, supported by high-capacity machinery, is particularly well-suited for large-scale seed hemp cultivation.

Despite its advantages, several limitations exist regarding the implementation of this method. These include a high risk of seed loss in the event of suboptimal system design or improper header selection. Furthermore, the mechanical wear and tear on machine components is significantly accelerated due to the abrasive and resilient nature of hemp stalks. Another challenge lies in the necessity of separating plant fractions deposited in windrows when they originate from different harvesting units; this, in turn, necessitates the integration of additional distribution or sorting mechanisms within the harvester’s architecture [11,22,28].

3.3. Selective Hemp Harvesting for Seed Production

Obtaining high-quality industrial hemp seeds while minimizing losses is a challenging task. Hemp plants ripen unevenly, and there are significant variations in plant height among different cultivars. Threshing entire plants—or even segmented sections—is complicated by the ease with which fibers are released, subsequently leading to clogging and blockages within the threshing and cleaning units. The previous section described machinery and harvesting methods that separate the seed-bearing parts from the stalks. These technologies mitigate the impact of the aforementioned factors on the efficient acquisition of high-quality seeds. However, more specialized machines designed for targeted, selective seed harvesting are also recognized and utilized in practical applications.

One such specialized machine is the B-800 hemp panicle mower (Figure 13), designed specifically for harvesting from seed plantations. This machine addresses a significant technological gap regarding equipment suitable for seed harvesting on small-scale hemp plantations. The mower was developed through a collaborative effort between the INFMP-NRI and the company FUGOR Sp. z o.o. (Krotoszyn), based on the conceptual designs and extensive research experience of the Institute [25].

The mower consists of two main parts: a cutting section with belt conveyors mounted on a front loader and a hydraulic power unit with an oil reservoir mounted on the tractor’s rear three-point linkage. The cutting assembly is a sickle bar consisting of alternating cutting knives and double separating guards. The cutting width is 2.5 m. Above the cutting assembly, there is a five-bat reel with an adjustable operating height. Mounting the mower on the tractor’s front loader allows for mowing within a range of 0.6 to 4.0 m above the ground. The drive of the machine’s working units is hydraulic, powered by the hydraulic unit. The control of operational movements is performed remotely via a control unit. The cut panicles are transferred onto belt conveyors, the second of which, featuring an adjustable operating angle, transports them to a trailer moving parallel to the mower. The minimum tractor power required to operate the mower is 102.9 kW [25]. As in the case of machines for harvesting hemp in longer segments, threshing the collected panicles is a separate operation, performed stationarily, most often using a threshing machine. Frequently, grain combines with a drum-type rasp-bar threshing unit operating in a stationary mode are also utilized for this purpose.

In both scenarios, the decoupling of threshing from the primary harvest—similar to other methods described where direct on-board threshing is not performed—generates additional costs. Beyond the operational costs of the stationary threshing equipment, these include expenses for material transport from the field, potential mechanical drying, and labor at each of the stage. On the positive side, this approach results in the low seed losses. According to research by Baraniecki et al. [25], for an average seed yield on flat terrain of 1540 kg·ha⁻¹, these losses amounted to approximately 2 kg ha⁻¹ DM. This is low compared to conventional combine harvesting, where seed losses are usually in the range of 46 to 26% of the harvested yield. This wide and imprecise range is due to the fact that, as with all machines and technologies discussed in this paper, efficiency is highly dependent on many environmental and biological parameters, such as maintaining the correct agrotechnical harvest date, weather conditions during harvesting, the specific hemp cultivars and the density of the panicles harvested for seed.

A machine with a comparable operating principle for the selective harvesting of hemp panicles is offered by another Polish company, Afori Sp. z o.o., based in Katowice (Figure 14) [57].

This machine is mounted on the rear three-point linkage of an agricultural tractor with a minimum power output of 90 kW (120 HP). The cutting height adjustment enables harvesting within a range of 0.8 m to 4.0 m. The operational efficiency of this mower reaches up to 10 hectares per working day. Once collected and potentially dried further, the panicles must be threshed stationarily as part of a separate procedure.

The previously mentioned company, HANF FARM GmbH from Malz, Germany, also offers its own version of a self-propelled harvester for the selective collection of hemp panicles. This machine is designated as the MultiCombine HC 3400 (Figure 15 and Figure 16) [58,59].

The machine is powered by a Deutz engine. It features a standard 4.5 m wide header with infinitely variable (stepless) height adjustment reaching up to 3.4 m. In its base configuration, the cutting unit is equipped with a double-blade cutter bar. The header utilizes a cross-belt conveyor instead of a traditional auger; its function is to transport the cut panicles to the right onto a second longitudinal belt conveyor, which subsequently transfers the material into a storage tank with a capacity of 20 m³. The MultiCombine HC 3400 is also designed to accommodate an interchangeable Shelbourne stripper header (Figure 16). For transport on public roads, the header is detached and placed on a header trailer towed behind the machine, following the same procedure used for conventional grain combines with standard harvesting units [61].

The harvester was designed with a high ground clearance of 1.6 m, which allows it to drive over the stalks remaining in the field after their tops have been severed. The machine can operate at a working speed ranging from 10 to 12 km·h⁻¹.

The stalks remaining in the field after the panicles have been harvested continue to stand (and can continue their growth cycle) until they are collected in a subsequent agrotechnical operation; for instance, using forage harvesters specifically adapted for hemp. Several such machines have already been discussed in this work [11,22,51,52].

4. Discussion

Implementing this method allows for seed recovery while significantly limiting the losses associated with the various operations required by previously described methods—with the exception of simultaneous panicle harvesting and segmented stalk cutting, or combine harvesting integrated with threshing (for low-growing varieties). By utilizing this harvesting technique, producers enhance both seed quality and homogeneity. This approach also reduces the moisture content of the material delivered for threshing in the subsequent on-farm operation. Harvesting panicles as they approach the end of their ripening period minimizes losses due to seed shattering. Even harvesting at full maturity is characterized by lower losses because the entire panicles are collected in the machine’s sealed tanks. In this scenario, any seeds that do shatter are also captured within these tanks, from which they can be safely recovered. In contrast, mowing plants and leaving panicles on the field surface consistently results in increased losses due to seeds shattering during subsequent necessary operations that cannot be performed directly on the field surface, as well as higher moisture levels resulting from contact with the soil [22].

Research Gaps and Technological Needs

An analysis of the extensive and diverse source materials presented above, combined with the authors’ own environmental, research, and engineering experience, has led to the identification of the following research gaps regarding effective harvesting technologies for industrial hemp, with a specific focus on seed production:

• Quantitative Seed Loss Models: There is a lack of a reliable, physics-based model capable of predicting seed loss as a function of panicle morphology, header dynamics, and environmental conditions. The development of such a model would be instrumental in designing optimal header speeds and integrated machine settings. To date, no such models have been developed due to the multifaceted impact of environmental and technical parameters. The magnitude of seed loss, as described in Section 3.3, depends on the variety, maturity and harvesting technology itself.
• Standardized Harvesting Unit Testing Protocols: There is a distinct lack of independent, comparative testing of both newly designed and currently operational harvesting units under standardized field conditions. The establishment of such a unified methodology would facilitate the accelerated implementation of best practices across the industry [62]. A review of the technology and the machines that implement it clearly indicates three possible methods of harvesting hemp. While many machines are available, most are single-unit productions or limited-run prototypes, or possibly small prototypes. The review also revealed that some of the machines, despite sharing the same general design and operating concept, are constructed from different components. This applies to both their chassis and transport units (e.g., a two-row machine on an Agrifac combine harvester chassis) as well as harvesting units themselves.
• Adaptive Combine Control Systems: The implementation of closed-loop control—utilizing various sensors to monitor both the execution of technological processes and the physical properties of the processed material—could significantly reduce seed damage and losses in real-time [21,22]. The review highlighted the lack of electronic monitoring and control over the harvesting, sectioning and threshing processes of hemp plants in many machines. Currently, such advanced systems are primarily found in modern, high-performance harvesters developed on established platforms (such as John Deere, New Holland, Kemper, or Claas), which are economically optimized for large-scale, uniform cultivation. Consequently, many custom-built machines require retrofitting with control systems based on precision farming principles. While the equipment level in Polish panicle harvesters developed by Afori and Fugor aligns with current technological standards, the integration of such systems into prototypes developed, for example, in Romania or China may prove technically challenging and cost-prohibitive.
• Application of Wear-Resistant Materials and Machine Design Based on Agronomic Principles: In the machine design process, careful attention must be paid to the selection of appropriate, durable, and abrasion-resistant materials, as well as to the environmental, biological, and agronomic factors stemming from the nature and morphological structure of hemp and the agrotechnical practices used for its cultivation (e.g., row spacing, plant density, maturity uniformity, plant height, hardness, elasticity). These aspects are discussed in the second part of the paper. In the case of sectional harvesting, the height of the plants is of utmost importance. This method, as well as dual-header systems, cannot be effectively utilized for cultivars attaining a height of only 1.0–1.5 m. For such heights, the former would essentially become a whole-plant harvesting method, while the latter would be technically unfeasible, as the dimensions and range of movement of both headers would not allow for effective operation below the minimum working height of the upper unit. Implementing these design considerations will undoubtedly facilitate better adaptation of machinery to specific cultivation systems, thereby optimizing the harvesting process, reducing machine downtime, and lowering associated operating costs. A comparison of the operating costs and area performance of some of the discussed machines is included in previous works [22,28,31,37,38,61].

5. Conclusions

Harvesting hemp for seeds remains more technically demanding than cereals harvesting due to the physiology of the panicle and the fibrous nature of the plant. Even the utilization of a conventional grain combine harvester for this purpose requires appropriate adjustments; e.g., by covering rotating parts that are not directly involved in the cutting, feeding and threshing process. The physiology of the plant necessitates a reduction in the rotational speed of the working units, which impacts work efficiency and economic indicators. Nevertheless, significant engineering progress has been made: dedicated harvesting units have been developed and machine settings and operating parameters have been modified to significantly mitigate seed losses and minimize operational disruptions during hemp harvest.

The most important factor in this regard is the selection of a harvesting technology appropriate for the crop. Specifically, the biological height of the plants is of paramount importance. For short-statured plants, it is advisable to harvest the entire plant, or utilize a combine harvester for simultaneous cutting and threshing; however, one must consider the potential of reduced seed quality due to the asynchronous maturation. This unevenness can be mitigated through drying, either on swaths or, more effectively, within industrial drying systems. In this case, Fugor and Afori panicle harvesting machines represent viable solutions. For the production of high-quality seeds in non-debarking hemp systems, the best results are achieved using methods based on selective panicle harvesting, whether as the primary product or through fractionated harvesting stages. In these cases, it is recommended to utilize machines equipped with dual harvesting units to facilitate the collection of panicles for seeds and the simultaneous cutting of the stems for field-drying in a single pass operation.

For dual-purpose (seed and fiber) hemp production, all the discussed methods are generally applicable. In the case of high-growth cultivars, producers without significant funds should prioritize harvesters capable of whole-plant cutting or systems that fractionate the plant into two or three segments. Alternatively, selective harvesting represents a more efficient yet capital-intensive option; consequently, this method is recommended primarily for large-scale industrial operations. For small-scale plantations, prototypes developed Romania, China, and Poland could be viable, provided they undergo further technical refinement and achieve commercial availability.

The hemp harvesting technologies from various countries discussed in this study, with a focus on seed production, demonstrate a broad spectrum of technical and technological capabilities. Both the innovative harvesting methods and the machinery designed to implement them are increasingly tailored to the specific operational needs of growers and the unique biological characteristics of the hemp cultivars they produce.

Furthermore, the research gaps identified through this comprehensive review of industrial hemp seed harvesting technologies have been clearly outlined, alongside recommended strategic directions for researchers and engineers to follow. Addressing these gaps is essential for the continued evolution of the industry and the optimization of harvesting efficiency.

The most important of these priorities are the development of quantitative seed loss models and protocols for testing harvesting units. Achieving this, however, necessitates the standardization of machine designs and the transition of current prototypes into their finalized, commercial-ready versions.