Hello Naysayers….. 😁
When I published a list of applications of Deep learning in Supply Chain a couple of months ago, many reached out to me saying that some of those may be just plain theory. The fact however was, as indicated in one of my previous posts, that I had tested every algorithm with Industrial scale data in last few months. Spent thousands from my own pocket to buy hardware to test some of these solutions.
One of the solutions in my article was- Trailer Utilization (Blog post link here: A Supply Chain Executive’s summary of Deep Learning : With 15+ innovative application opportunities across Supply Chain)
…..Say Hello to COROS (Cargo recognition and Organization system) from Daimler
Thanks to COROS solution from Mercedes Benz (Cargo Recognition and Organization system), developed for its sprinter vans. As per MB, the system helps optimize load optimization (make the best use of available space in the van), provides exponentially greater inventory tracking transparency and reduces expenditures related to training, shrinkage (due to lost packages) etc.
AI in last mile transportation has taken a big leap. The age old problem of load optimization, traditionally considered “NP Hard” by mathematicians, will soon have a non Knapsack type Industrial scale solution.
How does it work ?
The cargo sections of the van are equipped with cameras that can read bar codes from several feet away, using machine-learning algorithms to predict what that bar code is if it is obstructed or damaged.
When an employee is loading the van, the lights in the van will turn Red if the package does not belong to that van. If the package does belong to the van, a light will focus on the section in the van where it will be placed.
Why this is a very positive development?
My solution IS slightly different, in that it is designed for rackless Semi trailers, so the light guided architecture is not there but a different methodology will be used to guide the loaders on how to load. I won’t disclose what my solution leverages but as far as benefits of such a solution goes, it extends well beyond transportation.
I see two key additional benefits in the area of Warehousing and Inventory Management. Having visibility into packages means having an accurate data on how much Inventory is in the transportation system, where the inventory is at and when will the inventory cease to be our inventory.
Smart warehousing algorithms are my core area of interest and the data generated from this Deep leaning solution will be immensely useful to plan the optimal loading process as well, in addition to warehouse inventory tracking, flow and time standards planning and monitoring.
Exciting times ahead!
Congrats to the team at Daimler…& Thank You for proving me right !😁