At HighJump’s Elevate show in March, one of their customers – Hollar – spoke about implementing an autonomous mobile robot (AMR) solution and getting an almost instantaneous payback. I have been covering supply chain solutions for a long time and I’ve never heard that claim made. Really excellent payback can occur in six months, a two-year payback is about average, and some heavy forms of material handling pay back in four or five years.
The AMR solution was from InVia Robotics, and Hollar had also implemented HighJump’s warehouse management system (WMS).
AMRs from InVia Robotics were implemented at Hollar
Hollar is online retailer for products with prices starting at $1. They stock thousands of items from toys and electronics to home, beauty and apparel. The average order contains 12 items. The presentation at Elevate was given by Octavio Reyes, Hollar’s Director of Operations and Systems.
InVia Robotics is a Los Angeles based start-up company offering a robotics solution that navigates to an item’s location using QR codes, picks the desired item/s or container from standard warehouse shelving, then transports them to the packaging station or the next step in the fulfillment process. InVia’s robot, including a picking arm, is proprietary. The system includes a mobile robotic platform with a scissor lift that expands to eight feet in height, a vacuum grabber on an extendable arm, vision technology, and the software intelligence that manages the fleet of robots and its operations.
At Hollar, the suction device on the AMR grabs a paperboard container containing multiple units of a single product. The robot then brings that container to workers at a put wall station. Pickers pick the right number of items out of the cardboard container and puts the items in the correct plastic totes, the order totes, in a put wall. A video of the process can be viewed HERE.
HighJump Systems is a leading provider of warehouse management systems (WMS). Hollar started using the HighJump WMS in August 2017. InVia’s robots were deployed at the beginning of 2018. While the AMRs do most of the picking, some products are outside of the weight range that can be picked by InVia’s solution and the WMs manages those picks. The WMS also manages receiving, put-away, cycle counting, packing, and staging processes.
Mr. Reyes also pointed out that at the time they implemented the AMRs, InVia was a young company and there was a degree of risk the solution might not work. They needed the HighJump WMS as a backup to insure picking could proceed if the AMRs did not live up to their promise.
InVia’s robots are used in a goods to man process. Prior to this they were using a much more manual pick to cart process. The WMS also allows productivity to be measured. Without this baseline, they would not have been able to measure the increase in productivity of the new AMR system.
There were a few main drivers for implementing AMRs. As an online dollar store, they sell low cost, low margin goods. They needed a solution with the lowest cost per touch.
Secondly, like many companies, they have had a hard time hiring and retaining warehouse labor. This was particularly true as they were growing, they have big seasonal surges in sales, and the temp workers hired were not as productive as their existing workforce. Out of the one thousand or so potential employees that walked through their doors, they only ended up with 35 to 40 good employees. The AMRs enhance productivity. This meant that Hollar did not have to hire as many employees. These AMRs increase the number of orders that can be picked by a factor of five, so Hollar can operate with one fifth of the work force they would otherwise need.
Finally, because Hollar was growing, they knew they would need move their warehouse to a more central location in the US (they ended up selecting Cincinnati). They wanted flexible automation that would be quick to implement at the new site.
So how can the payback be almost immediate? The InVia solution is sold in a Robots as a Service (RaaS) transactional model. InVia charges a low price for each robot cycle – picking the paperboard carton, bringing it to the pack wall, and returning the carton to the shelf. Under RaaS pricing “If the cost for the robot to do these activities is less than the cost of a human, than the payback is immediate.”
Well not quite immediate. Hollar still had to pay for the implementation. It took 3 weeks to design the solution and two weeks to implement it. That is fast. So, while the cost of the implementation was not mentioned, it sounds much less expensive than most other automation solutions.
Here were a few other points I found interesting in the presentation:
- These AMRs need smooth floors to operate efficiently.
- A special tape was put on the cardboard boxes to improve the ability of the suction system to pick items.
- The current maximum weight that can be picked is 50 pounds.
- It is rare for the robot to not be able to pick an item. There are only one- or two-times Mr. Reyes remembers that happening. In both cases, it was because the case weighed more than 50 pounds and the system was not aware of that fact.
- The AMRs contain their own optimization logic. It can be different from the logic you would use for humans. For example, with humans fast moving goods should be put away in the golden zone, a zone that does not require bending or stretching. For the AMRs, the golden zone is close to the floor so that the scissor lift does not have to be deployed.
- Training new workers use to take 2 to 4 hours. In the goods to man process facilitated by the AMRs, new workers only need 30 minutes of training.
In conclusion, the AMR market is the fastest growing market I have ever encountered. It is easy to see why.