MHI Solutions Improving Supply Chain Performance
Higher Education
What does the ergonomic data tell us about these low-tech but essential selector tools?
Figure 1. Case hook being used to access a second level case.
Full case selection processes in distribution centers (DCs) frequently have selectors obtaining products from two level racks. Given the pallets are usually stored with the 40-inch side facing the aisle, this makes reaching to the back half of the 48-inch deep pallet to obtain products physically challenging. This is particularly true when selecting from the second racking level, where bottom layer products may start at 50 to 55 inches above the floor and be stacked as high as 84 inches above the floor.
DCs address this challenge by providing the selectors with “case hooks” (a.k.a. “pick hooks”) so they can more easily access the cases of products located further back on the pallet (Figure 1). In a prior study on ergonomics practices in manual selection operations (Lavender & Sommerich, 2022), we found that every DC we spoke to had some sort of case hook that was available to their employees. However, utilization varied considerably across facilities, as did the case hook designs. Therefore, we sought to identify case hook designs that can effectively reduce the biomechanical loads during 2-tier case picking operations, and are perceived by users to be useful, usable and desirable.
Prior work using focus groups with grocery selectors at nearby distribution centers identified different design features that were worthy of biomechanical investigation (Figure 2). Based on these design concepts, prototypes were developed and tested in a laboratory environment to determine if different handle and tip designs led to lower biomechanical demands during a simulated case pulling process.
Figure 2. Design attributes from focus group sessions with DC workers.
Our laboratory study engaged 12 volunteers, ten males and two females, that ranged in height from 5 foot 5 inches to 6 foot 2 inches. These research participants performed a simulated product section tasks using four different handle designs and three different tip designs (Figures 3 and 4), for a total of 12 tool design variations. The handle designs included an in-line handle that was a straight tube, a T-handle, a pistol grip handle and a counter-balanced pistol grip (Pistol-CB) that had a weight attached to the proximal end to bring the tool’s center of gravity to the pistol grip handle. The tip designs included a conventional tip that was a narrow flange orthogonal to the pipe, a large hook that was designed to facilitate reaching around the side of the box being selected, and the rake, which was a flat blade surface that was orthogonal to the pipe. All tools were made from PVC material, were 85 cm in length from the center of the grip to the tip, and weighed between 1 and 1.6 pounds, except for the counterbalanced versions, which ranged from 3 to 3.5 pounds.
The participants were asked to pull 12-pound boxes forward on the pallet to where they could easily be grasped for palletizing. This weight was selected as it is representative of case weights frequently slotted within second level slots in grocery distribution operations. While the participants were pulling the boxes forward, the electromyographic muscular activity (EMG) levels in the shoulder and wrist muscles were sampled, as well as their shoulder postures using a motion capture system.
The sequence of handle design and tip conditions was randomized for each participant. Within each condition, there were two trials. At the completion of the session, the participants were asked to rate the usability of each handle and tip configuration on a 5 point rating scale (very difficult to use to very easy to use).
Research Results
The analysis of the kinematic data showed that both the type of handle and the tip design significantly (p<.001) affected the degree to which the shoulder swung forward (shoulder flexion) and to the side (shoulder abduction). Figure 5 shows that pistol grip resulted in reduced motions in both planes relative to the other handle designs tested. This figure also shows that within the pistol grip designs were only minimal differences due to tip design. It should also be noted that when the box was further forward on the pallet, the shoulder flexion and abduction decreased for the pistol grip handles, but not for the Inline or T-handle.
Figure 6 shows the pistol grip led to lower muscle activations (90th percentile normalized EMG) for the anterior and lateral parts of the deltoid muscle, as well as one of the key muscles controlling wrist movement in this task (flexor carpi radialis). Looking at the pistol grip data more closely, it can be seen that for the shoulder muscles, there were no significant differences in the results due to tip design. For the tested wrist muscle, the hook tip led to significantly greater muscle use in both pistol grip designs and with the T-handle.
Usability data also showed statistically significant effects (p<.01) for both the handle and the tip designs. Figure 7 shows the usability ratings for the handle and tip designs, where higher values indicated a design that was easier to use. These data suggest the inline and pistol grip handles were perceived as easier to use than the T-handle. The rake-style tip was also rated easiest to use.
Both the analysis of shoulder postures and muscle activation data suggest the use of pistol grip tools to reduce the biomechanical loading of the shoulders when using case hooks to obtain products. These findings are also supported by the usability data wherein the pistol grip was considered one of the easier tool handles to use. This may be due to pistol grip affording information regarding the tip orientation. While the same could be said for the T-handle, extra effort was required to control the pitch of the tool as the handle could more easily rotate within the hand.