Vertical lift module throughput is rated in trays per hour (TPH) or order lines per hour (LPH), and the 2026 vendor range spans roughly 100-300+ LPH for a single unit, with 300+ LPH achievable only on dual-elevator, double-deep, or multi-machine podded setups [S1][S3][S6]. A standard single-elevator vertical lift module typically delivers 100-120 tray presentations per hour, while a dual-staged elevator such as the White Systems Compact Twin pushes past 100 TPH by staging the next tray while the operator picks from the current one [S1].
The honest ceiling is mechanical, not marketing: a single extractor on a single lift shaft interrupts picking every cycle because the operator cannot reach the next tray until the current cycle completes. Pods of three or four VLMs working in parallel can hit several hundred lines per hour, because the operator always has a tray waiting at the access window [S7]. Anyone specifying a VLM on a published "300+ LPH" headline without checking the configuration context is going to oversize or undersize the install.
Where the 100-300+ LPH Number Actually Comes From
White Systems' April 2026 guide states the Compact Twin dual-elevator VLM achieves "over 100 tray presentations per hour," described as "nearly double the throughput of a standard VLM" [S1]. Modula's product literature for the same year quotes "up to 300+ lines per hour" for its VLM line, with a separate spec sheet pinning the tray-presentation number at "up to 120 trays/hr, depending on configuration" [S3][S6]. The 300+ figure is the order-line rate an operator can complete when the machine is delivering trays faster than the operator can pick them; the 120 trays/hr figure is the machine's intrinsic delivery rate, which is the binding constraint [S3].
Academic throughput modeling (Vanhauwermeiren et al., 2020) parameterizes TPH as a function of lift vertical speed, extractor horizontal speed, acceleration, tray-transfer time, and VLM height, and concludes that the dominant lever is not raw lift speed but the effective availability of trays at the picking bay [S4]. In practice, that means a 2 m/s lift on a 10 m unit will not double throughput over a 1 m/s lift on a 5 m unit if both are running single-elevator, single-extractor configurations, because the operator's pick time still gates the cycle. The peer-reviewed model is the cleanest evidence that vendor "X LPH" claims must be read against the operator pick time, not in isolation.
Single, Dual, Double-Deep, and Podded Compared
Four VLM configurations dominate the 2026 spec conversation, and they line up against four decision criteria: nominal LPH, peak LPH with a fast operator, capital cost per LPH, and best-fit order profile. The comparison is built from named OEM data plus the parallel-machine reasoning in the Kardex design guide [S1][S2][S3][S7].
Single-elevator / single-extractor (Compact Lift class): 60-100 TPH nominal, 100-150 LPH with a skilled operator, lowest cost per LPH, best for labs, MRO, and slow-moving spares at 10-50 order lines per day [S1][S2]. Dual-elevator (Compact Twin class): 100+ TPH nominal, 150-200 LPH achievable, mid-cost, best for mid-volume parts distribution and batch picking where the elevator can stage the next tray during the current pick [S1]. Double-deep (Compact Double class): storage density up 85-90% versus standard racks, but throughput per machine is comparable to single-elevator because both elevators still feed one picking bay, so the trade is space-per-LPH rather than LPH per dollar [S1][S3]. Pod of 3-4 VLMs in parallel: 250-400+ LPH at the workstation, highest cost and footprint, best for e-commerce DCs running hundreds to thousands of order lines per hour where single-machine cycle time is the binding constraint [S2][S7].
The key decision rule from the Kardex sizing guide: if operators are standing and waiting for the next tray, the problem is machine count, not machine speed. Throughput-led designs should split inventory across two, three, or more VLMs and connect them with WMS-level batch logic so several machines move at once [S2]. A single taller machine, even at 50+ feet of clear height, will not deliver high LPH because the elevator travel and the single picking bay are still serial [S2][S5].
Use-Case Fit: Who VLMs Are For and Who They Are Not
VLMs are for facilities that need to consolidate slow-to-medium-velocity SKUs into a small footprint and are willing to accept goods-to-person pick rates in the 100-300 LPH range per workstation. The Kardex guide identifies labs, R&D rooms, and small parts operations at 10-12 accesses per day as the sweet spot for one tall machine, and high-speed distribution centers processing hundreds to thousands of order lines per hour as the sweet spot for multiple machines in parallel [S2]. The Modula application list extends to medical, aerospace, automotive, and electronics small-parts storage where ergonomic tray presentation at the "golden zone" (37-40 inches above the floor) reduces bending and ladder use [S1][S3].
VLMs are not for full-pallet flow, bulk storage, or operations where the pick face must move with the picker. A VLM tray is generally 30-40 inches deep and up to 12-13 feet wide, so any SKU that exceeds the largest tray or the per-tray weight limit (Modula: 2,182 lb/tray; Kardex Shuttle 1000: 2,204 lb/tray) belongs on pallet racking or in a mini-load AS/RS instead [S3][S5]. Facilities that need true conveyor-integrated sortation at 1,000+ units per hour should evaluate horizontal carousels in pods, mini-load, or unit-load AS/RS, where the cycle-time economics are different from a single VLM bay [S7].
Engineering Levers That Move the LPH Number
Five variables move a VLM's real-world lines-per-hour figure, and OEM brochures vary which ones they disclose. Vertical zoning keeps fast movers in the lower trays so elevator travel stays short, which is the single most effective non-hardware lever [S2]. Lift vertical speed and acceleration determine the floor of the cycle time; the 2020 throughput model shows that doubling lift speed on a tall unit produces less than a 2x gain in TPH because operator pick time still gates the cycle [S4]. Extractor (horizontal tray-transfer) speed is usually faster than lift speed, so it rarely bottlenecks unless trays are oversized.
Drive architecture is a reliability variable, not a throughput variable: rack-and-pinion elevators (used in White Systems units) do not stretch over time the way belt or chain drives can, which keeps cycle time consistent over a 24/7 duty cycle [S1]. Tray height auto-measurement (laser on putaway) eliminates "honeycombing" wasted vertical space and indirectly raises effective storage density by 3-4x versus standard racks within the same footprint, which means more picks per square foot without changing the machine [S3][S5]. The published Modula spec of "30 minutes to train an operator" is also a throughput lever: the faster a new picker reaches steady-state pick time, the sooner the LPH number on the brochure matches the LPH number on the floor [S3].
Limitations, Failure Modes, and the Operator Bottleneck
The fundamental VLM throughput limitation is the single picking bay. The lift can move faster, the extractor can move faster, but the operator's hands and barcode scanner set the upper bound on a single-machine install. The Battini et al. (2016) study of dual-tray VLM batch picking measured a picking time per line of 6.1 seconds in a six-customer batch configuration, which translates to roughly 590 LPH at the pick-action level, but that figure assumes batched orders, ideal SKU adjacency, and uninterrupted tray flow [S9]. Real single-machine installations rarely sustain that because tray changeover and putaway eat into the same cycle.
Common failure modes that quietly crush the rated LPH include: mis-sized tray inventory so the elevator hunts for the right slot height, slow pick-to-light or scan workflows at the access window, software that does not pre-stage the next tray while the current pick is open, and per-tray weight or dimension limits that force split picks across multiple trays [S2][S4]. Vertical lift modules also do not scale linearly: doubling the height of a unit does not double throughput, because the elevator's round-trip travel grows with height while operator pick time stays constant [S4]. This is why the 2026 design advice from multiple OEMs converges on adding machines in parallel rather than building one taller, faster unit [S2][S7].
Standards, Sourcing, and What to Confirm on the Datasheet
VLM performance is not governed by a single ISO or IEC product standard the way pressure transmitters or flow meters are, so verification falls on the datasheet and the OEM's documented test conditions. The verifiable items on a 2026 VLM datasheet are: nominal TPH under a defined operator pick time (e.g., 5 s, 8 s, 10 s per pick), maximum tray dimensions in inches or millimeters, per-tray weight in pounds or kilograms, unit clear height in feet or meters, drive type (belt, chain, or rack-and-pinion), and controller/WMS interface protocol. Modula's published 2,182 lb/tray (about 990 kg) and Kardex's 2,204 lb/tray (1,000 kg) on the Shuttle 1000 are the upper weight bounds a specifier should anchor against [S2][S3].
For facilities integrating a VLM into a broader WMS or WCS stack, confirm the API or middleware support, batch-pick logic, and pick-to-light or scanner hand-off at the access window before signing the PO, because software latency is often the gap between "300+ LPH on the brochure" and "180 LPH on the floor" [S2][S3]. The 2026 Modula spec sheet already integrates WMS inventory control as a standard feature, which is the bar to set when evaluating alternative VLM brands [S3]. Related reading on footprint trade-offs is in the VLM vs Floor Storage: Footprint, Density, and Throughput Compared reference, and the install-side constraints (floor load, clear height) are detailed in VLM Installation: Floor Load and Ceiling Height Specs. For operations pairing VLMs with manual assembly or kitting cells, ergonomic loading limits and reach distances covered in Industrial Exoskeletons on Assembly Lines: Injury Reduction Evidence, Limits, and Sizing are the adjacent spec conversation.
Trackable signals to watch over the next two quarters: OEM disclosures of TPH under a stated operator pick time (so the 100 vs 300 LPH number can be compared apples-to-apples), the first published multi-machine pod throughput case studies with batch-pick software, and any WMS-vendor integration announcements that close the pick-action latency gap. A 2026 specifier should also pin the test condition in writing, because the difference between a 100 LPH and a 300 LPH VLM is almost always configuration, not the machine.
The underlying component specifications are covered under linear module, and relay module.