FEM 9.851 defines the performance data and six canonical cycle cases for storage and retrieval machines, distinguishing three load-handling types, with the double cycle (Doppelspiel) the canonical throughput metric reported in DC/h [S1].
Cycle time is the sum of constant and variable time components: constant parts are positioning, switching and brake delays (a typical t_tot around 1 s) plus the LHD store/retrieve time, while variable parts come from the travel, hoist and LHD motion profiles for each segment E-P1, P1-P2 and P2-A [S1].
The six cases and three load-handling types in FEM 9.851
FEM 9.851 is a non-profit trade-association standard that codifies six storage and retrieval cycle cases and three load-handling types (double cycle, single storage, single retrieve), with Case 1 in the standard being the reference double-cycle example [S1]. A double cycle executes one storage and one retrieval in a single crane movement, so the headline throughput metric is double cycles per hour (DC/h) rather than single cycles, and a 1 DC/h figure equals one inbound plus one outbound pallet handled per hour by one machine [S1]. In Spanish-speaking markets the equivalent reference is UNE 58912, which lists analogous movement scenarios for combined cycles in single-deep pallet and box AS/RS [S2].
For a stacker crane the kinematic chain that drives every case is the same: travel along the aisle (x), hoist along the mast (y) and the load handling device stroke (z), with FEM 9.851 requiring the specifier to declare masses, peak and effective acceleration, peak and effective emergency-stop deceleration, and the velocity profile (trapezoidal, jerk-limited, or rounding-time) for each axis [S1].
Building the cycle equation from kinematics and distances
Each segment t(E;P1), t(P1;P2) and t(P2;A) is computed from the axis velocities, accelerations, profile type, and the straight-line distance between the two points, with a separate hoist contribution added whenever the destination level differs from the source level [S1]. When the entry point E and the exit point A are not the same conveyor, the cycle equation also needs the return travel from A back to E, which is why the total double-cycle time is usually written t_DS = t(E;P1) + t(P1;P2) + t(P2;A) + t_uber + t_tot, with t_uber the LHD store/retrieve time and t_tot the constant positioning/switching/brake delay [S1][S2].
Industry guidance for intermittent motion equipment follows the same logic, since duty cycle math for intermittent linear actuator operation treats motion profiles as a sum of constant plus variable components, which is the pattern the FEM 9.851 segment equations follow. Specifiers should also confirm that the declared velocities are the rated steady-state values, not peak motor speeds, and that effective acceleration reflects average torque over the ramp, not the motor's short-duration peak.
Reference cycles for mean energy and throughput comparisons
Energy demand of a stacker crane cannot be averaged across arbitrary operation patterns, so FEM-aligned researchers define a small set of reference cycles that represent typical AS/RS workloads and let the mean energy demand be back-calculated from those reference cases [S3]. The TUM fml study distinguishes miniload stacker cranes (MSC, for small load carriers) and pallet stacker cranes (PSC, for unit loads) because the mass ratios between payload, mast and counterweight dominate the regen-braking share and therefore the net energy per cycle [S3]. The same reference-cycle approach underlies FEM 9.841-era throughput models, where travel time is computed from a stochastic analytical travel-distance model and then the cycle time is closed with the LHD handling time, which is exactly how Siemens' stacker-crane inquiry data sheet asks the buyer to enter DC/h, payload handling time in seconds, and the per-axis kinematics [S1][S4].
Mean energy per double cycle is sensitive to three parameters in roughly this order: payload-to-empty-mass ratio (dominant for PSC), vertical lift share of total motion (regen recovers more on descent), and the dwell time at the storage position because that dwell sets how much auxiliary load is drawn between productive moves [S3]. When two cranes are being compared for the same warehouse, the right comparison is energy per DC (kWh/DC) at the agreed reference cycle, not kWh/h at the maximum DC/h rating, since the maximum rating is rarely reached in a stochastic storage policy.
Two-step storage policy and dual-command cycle times
A two-step storage policy explicitly distinguishes a storage window (only inbound moves) from a retrieval window (only outbound moves), and within those windows single-command and dual-command cycles are computed separately, with the waiting time of a single-crane aisle entering as a penalty term when the storage and retrieval windows are not perfectly balanced [S6]. For a single-crane aisle the dual-command cycle time t_DC equals t(E;P1) + t(P1;P2) + t(P2;A) plus two LHD handling times, while the single-command cycle t_SC is t(E;P1) + t(P1;A) plus one LHD handling time, and the average cycle is a weighted mean of the two depending on the storage/retrieval mix [S6].
For two-crane aisles the same equations apply per crane but the waiting-time term drops, which is why high-throughput freezers and pharma AS/RS often specify twin-crane aisles even though the second crane sits idle for long stretches, since the throughput ceiling at the peak hour is what sets the warehouse capacity, not the average DC/h. The practical selection question is therefore: if the required DC/h is more than what one crane can deliver at the L_max and H_max of the building, twin cranes on a single rail become economic, while below that threshold a single crane with a slightly larger mast height is usually cheaper.
Selection criteria: cycle time, energy, and DC/h trade-offs
Four decision criteria separate the main options for specifying FEM 9.851 cycle time, and the right answer is a weighted mix, not a single number: cycle time (lower is better, dominated by axis velocities and effective acceleration), energy per double cycle (lower is better, dominated by mass ratio and regen share), DC/h throughput (higher is better, dominated by aisle length and lift height), and machine capital cost (lower is better, correlated with mast height, top speed and number of axes with servo control) [S1][S3]. A pallet stacker crane (PSC) optimized for high DC/h typically runs 2.0-3.0 m/s travel, 0.5-1.0 m/s hoist, and 0.3-0.5 m/s LHD with 0.5-1.0 m/s^2 effective acceleration on each axis, while a miniload stacker crane (MSC) for small load carriers usually runs faster travel (up to 4-6 m/s) with much lower hoist speeds because the bin weight is small, and the FEM 9.851 inquiry sheet captures exactly this by separating the velocity and acceleration fields for travel, hoist and LHD, with payload and empty values where they differ [S1][S3].
For a typical 50 m aisle, 10 m lift height pallet AS/RS running a balanced 50/50 storage and retrieval mix, a single PSC with 2.5 m/s travel and 1.0 m/s hoist will land near 60-80 DC/h, while the same aisle served by an MSC for 50 kg bins can exceed 120 DC/h because the hoist time scales with payload and the MSC payload is roughly an order of magnitude smaller, which is why the same FEM 9.851 case definition can give very different DC/h numbers depending on machine class.
Who FEM 9.851 is for, and where it stops
FEM 9.851 is the right tool when the question is how many DC/h a single storage and retrieval machine can deliver at a given aisle geometry and kinematic envelope, and it is the format almost every European OEM and system integrator asks for in the inquiry phase, including Siemens' standard inquiry data sheet for storage and retrieval machines [S1]. It is not a control-software standard and it does not cover warehouse management logic, slotting, or order sequencing, which is why the TUM reference-cycle work and the two-step storage policy literature treat the cycle math as one input to a larger simulation rather than a complete warehouse model [S3][S6].
The standard is also quiet on multi-crane deadlock resolution, fire-mode operation, and the cold-aisle derating of cycle time that is standard in freezer warehouses, so specifiers in those environments should layer those effects on top of the FEM 9.851 result rather than expecting the standard to deliver them. For throughput planning beyond the inquiry phase, FEM 9.851 outputs are normally fed into a discrete-event simulation of the full AS/RS, which is also how the stochastic analytical travel-distance model in the TUM work was validated against measured cycle times [S4].
Standards, sources, and what to track next
Trackable signals to watch on FEM 9.851 in the next planning cycle: any FEM technical-guidance update that revises the reference cycle set or the axis-profile taxonomy, which is published under the Intralogistic Systems product group of the FEM technical-guidance catalogue alongside documents such as the cycle-time guideline for automated vehicle storage and retrieval systems (2017) [S5]. The TUM fml reference-cycle paper, the UNE 58912 Spanish standard on combined cycles, and the two-step storage policy work in MDPI Processes (2021) are the three most cited open references for extending FEM 9.851 into energy and throughput comparisons [S2][S3][S6]. For spec writing, keep the FEM 9.851 segment equations, the Siemens inquiry-sheet field set, and the UNE 58912 combined-cycle case list on the same page so that any DC/h number is reproducible from the same kinematics block.
Component reference pages worth checking: time relay, and pallet stacker.