Distribution and A.I.

Machine Learning and A.I.  trained to watch distribution operations will revolutionize logistics 

I began running supply chain and logistics operations over 40 years ago.  In those days virtually all controls were paper-based with big computers limited to pushing materials onto the factory or distribution floor via the early versions of material requirement planning (MRP) systems.   The real management of distribution, including shipping, receiving, traffic, and warehousing,  was based on on-site supervisors and leads watching and directing the action.

In the following decades, bar cording, RFID, and mobile computing help improve the efficiency (speed) of moving and tracking materials from point to point.  Even with the improvements, on-site supervision was essential to keep things moving along, to flag safety violations, and discover lapses in training. With wave after wave of cost cutting, the numbers of on-site supervisors and team leaders has continued to decline.

While artificial intelligence and machine learning go back decades, it is only in the last few years that the cost and ease of acquiring and storing data has dramatically improved. It is now possible to apply computer vision to help distribution in a wide variety of areas:

  • Vehicle in and Out Cycle Times. Computer vision can watch and identity each vehicle entering and exiting a designated area reading license plates or other unique identifiers with time stamps for each transaction. The digitization and visualization of the process can include dashboards, a variety of alerts, and datamarts for more in-depth analysis. 
  • Speed Violations in the Yard. Computer vision can track the speed of vehicles in designated areas and send out alerts with a visual record of the violation. With dashboards, a daily and weekly analysis can reveal previously unavailable patterns of speeding and other violations.
  • People in the Wrong Areas.  With computer vision, it is easy to designate restricted areas in which no one should be on foot. It is also possible to identify types of people permitted in a designated area by the color of their safety vests or head gear.  This issue is especially important in mining, quarry, and landfill operations in which very large vehicles make it dangerous for anyone to be on foot.
  • Vehicles Parked in the Wrong Areas.  In large yards, it is not unusual for vehicles to get lost in the system.  As long as there are unique identifiers, computer vision can locate vehicles outsize of their designated areas. This can be an issue in areas with tight space limitations such as service vehicles around aircraft or trucks parked too close to earthmoving equipment.
  • Fork Lift Operators.  Fork lifts are synonymous  with warehousing and distribution but are inherently dangerous to operate.  Their efficient and safe use can be improved by visually monitoring vehicles and their operators.  Problems such as speeding, driving in restricted areas, or not wearing safety equipment can all be flagged and alerted in real time with a visual record and time stamp.
  • Meeting Critical Departure Times.  In many operations vehicle departure times must be closely coordinated with the loading times of other trucks, trains or aircraft. One of the most critical of these are departure times for catering trucks servicing wide-body commercial aircraft.  Delays of even a few minutes can create major collateral damage. With computer vision it is possible to capture the exact minute of departure and flag delays.  The alerts can be triggered by data feeds from other systems providing planned departure times. 

    

These are just a few examples of how computer vision is revolutionizing distribution. They come from Atollogy’s current clients in food processing, quarries, landfills, airports, and retail distribution centers. These clients are now asking us to explore additional areas in which to apply computer vision based on A.I. So, the list of applications is going to grow and grow quickly.

Author:

Anthony Tarantino, PhD

Adjunct Professor, Santa Clara University – Operations, Finance, Lean Six Sigma

Six Sigma Master Black Belt, Certified Scrum Master, CPIM (APICS), CPM (ISM)

Senior Advisor to Atollogy

Tony@atollogy.com