Automated Storage and Retrieval Systems: The Impact of Load Condition on ASRS Reliability and Availability

This is Episode 5 in a series of “Common Mistakes” and focuses on understanding importance of load condition as it relates to system performance. 

Understanding Reliability and Availability

When planning an automated storage and retrieval system, you’ll see claims such as: “guaranteed 99% reliability” and “guaranteed 99% availability”. Percentages may vary based on supplier and application. 

Reliability is a measurement of the probability that a requested operation will be successfully completed. It is calculated by dividing the number of successful operations by the number of attempted operations.

Reliability %   =            (Total Number of Successful Operations)

                                       (Total Number of Attempted Operations)

It is not unusual to see reliability guarantees of 99%.

Availability is a measurement of the probability of finding a piece of equipment, at any given time during the period of operation, in a state which will allow a requested operation to be carried out correctly and without malfunction. It is calculated by dividing uptime by operating time.

Availability %   =          (Total Expected Run Time) – (Internally Caused Downtime)

                                                                   (Total Expected Run Time)

It is not unusual to see availability guarantees of 99%.

A chargeable error condition is one due to equipment performance or failure. Excluded from reliability and availability calculations are faults and failures related to load condition. In a typical pallet ASRS, the majority of system faults are related to pallet/load condition, not equipment failure. Also excluded from availability calculations are the response times; the time from when the fault/failure occurred until the time when trained personnel arrive to address the error.

Using an example system, we can illustrate as follows:

• 100 loads in and 100 loads out per hour
• 20 hours per day, 5 days per week
• 10,000 loads in and 10,000 loads out per week
• 99% reliability means that 200 chargeable faults/failures could occur in a week
• 99% availability means there can be 1 hour of chargeable downtime per week

But, the scenario above is rarely reality.

Modern equipment is highly reliable --- equipment failures are quite rare. In fact, as experienced with systems handling totes or high quality pallets and loads, it is not unusual to see reliability rates that exceed 99.98% or 99.99%, including faults and failures of all types. In the example above with 20,000 load moves per week, that corresponds to 2 - 4 errors per week. Systems with good loads that are well suited to the equipment design, “run like a Swiss watch”.

On the other hand, there are many examples of systems where the pallet condition or load size and condition is poor and not well suited to the equipment design. Yes, the load may pass the size check station, but a pallet or load may degrade while being transported through the system. For example, a stretch wrap tail that was marginally affixed at the size check station may become dislodged in transport causing an equipment fault.

Playing out the previous scenario, we’ll pretend that 1 in every 200 loads are such that they cause an equipment fault or failure while be transported, stored, and retrieved --- or, 99.5% of the loads are good, and .5% of pallets or loads are sub-standard per the application (and generally not chargeable back to the system supplier).

Since automated equipment is natively highly reliable and most errors are related to load condition, the reliability of the system may still be relatively high --- greater than 99.4%. The true impact is the effect on system availability and the knock-on effect on system performance.   

In this example, ? percent sub-standard loads equates to 100 load related faults or failures per week. A typical fault or failure takes about 20 minutes for recovery time: fault occurs, personnel respond, rectify condition, operation resumes. 

20 minutes x 100 incidences = 2,000 minutes or 33 hours --- or, an availability of 67%! True, not every fault affects the entire system, but not every fault can be rectified in 20 minutes either. The impact on system availability and operational performance is dramatic.

The “Common Mistakes” are:

• Poor match of equipment design to the pallet, load, or carrier --- including load handler type, racking type, operating tolerances, and conveyance designs.
• Insufficient focus on high quality pallets and loads that don’t cause “load related errors”. A small percentage of loads can have an outsized negative impact on system performance.
• Designing systems such that rectifying a fault causes stoppage of large portions of the system. 

If you want your automated system to “run like a Swiss watch”, then be sure that equipment and system design takes into consideration the condition of your pallets, loads, or carriers.

Are you planning to build an advanced warehouse automation system? If so, there are several steps leading to a successful outcome: data analytics, simulation model, and an experienced system designer. Don’t start your project without them!