Case Study

Fleet Monitoring and Analysis

Now more than ever, it is critical for a modern warehouse to manage forklift efficiency by reducing cycle time and allocating forklifts to the right locations. Efficiency is essential in a large or urban warehouse where there are many vehicles, and the product turnover ratio is high. Motion2AI’s Monitoring Service provides metrics to support precise decision making by combining two sources of information:


1) The MotionKit, an easily attachable forklift module
2) WMS data from the client


This combination provides metrics—such as utilization, exact cycle time, and index—that improve put-away/vehicle-allocation strategies.


Company A, a large international logistics company in Korea, benefited from a three-month pilot with Motion2AI. They reduced the number of operating forklifts, and they revised the put-away strategy to alleviate bottlenecks. They optimized the routing algorithm, and they figured out the burdens of each job type.

Vehicle Utilization


Most warehouses depend only on their WMS data to determine the number of operating forklifts using cycle time, capacity, or turnover ratio. However, there are other variables like volume volatility, battery discharging, or different forklift type requirements. Therefore, the operating company tends to deploy more forklifts than they need, based on a conservative approach.


Company A operates 22 forklifts of 4 types: reach, order-picker, counterbalance, and side clamp. All forklifts had MotionKits installed, which send out location data every second. This information shows utilization for each vehicle (see table below).

Vehicle utilization is displayed by the hour when each vehicle was in use.


The figure above is a dashboard feature displaying the utilization of Company A’s 22 forklifts. It is easy to determine how many vehicles are used during peak times by coupling it with WMS data. Company A can remove 4 vehicles, an 18% reduction.

Bottlenecks


MotionKit location data show where vehicles are stationary for extended periods.



Red dots represent where vehicles are stalled for more extended periods. These areas can be bottlenecks of working processes.


Many forklifts were moving between floors through elevators in Company A. They spent significant time waiting for elevators. Company A will make a new put-away strategy to reduce these movements.

Working Times


Working hours by job type and area can be measured using only the information obtained from the MotionKits. By setting working areas on a map, the duration in that area can be calculated for each forklift.


Measuring working hours by setting areas for specific job types.


Company A could measure loading, unloading, and elevator-waiting time by setting corresponding areas on the map. Picking resulted in the most significant burden, followed by loading, unloading, and then put-away. Company A could prioritize their processes by load and optimize for picking jobs first.

Evaluating the Routing Algorithm


By plotting order routes on the map, we can evaluate the routing algorithm.



Picking order example.


There are several inefficiencies in Company A’s routing algorithm; some locations are ordered in a sub-optimal manner where the overall travel distance could be decreased. Company A will prepare for an optimized routing algorithm in the future.

Put-away Strategy


To optimize operations, ABC-analysis is required to place top SKUs in good locations. Unless the put-away strategy goes along with the ABC-analysis in real-time, there can be inefficiencies in the put-away strategy.



Picking frequency by rack. The darker the color, the more often it is visited.


Company A’s storage seems to be managed well, but several locations have not been used for a long time, near the loading area. By increasing the location turnover, Company A expects to reduce moving distances, including movements between floors.

Conclusions


Through the use of MotionKits and a monitoring system service deployed on the cloud, Company A achieved reduced operation costs, set up a real-time monitoring system, received vehicle utilization metrics, and analyzed routing and rack efficiencies. Since most other warehouses operate like Company A, we expect more opportunities to optimize for other clients.

 
 

Written by Sungho Kim, Data Scientist at Motion2AI.