Logistics Optimization and Artificial Intelligence Development
Can we challenge the current paradigms of planning and procuring transportation? How could it be modernised? How can we be more efficient – both in cost and for the sake of the environment?
Recently at Uptrend Labs, we have been thinking about this issue after reading on how inefficient and costly is shipments in the U.S.
For a long time, shippers have been the decision-makers when it came to modes of transport; predominantly whether to do a Full Truck Load (FTL), a Less-than Truck Load (LTL), Groupage, or a Parcel one.
Decisions are often (and unfortunately) based on rough estimations (e.g.: the size of shipment) and always centred on the shipper’s own optimisation methods and processes (e.g.: merging different customer deliveries into single movements based on proximity).
Even today, Shippers are making their choices based on three main factors, which have all become increasingly volatile and variable post Covid-19:
- Availability
- Lead-time
- Cost dimensions
Many Shippers try to engage 3rd Party Logistic providers (3PL’s) to reduce their costs, relying on them to take the burden of optimizing shipment requirements, service lead-times and still satisfy customers’ needs. On their side, 3PL’s can use a wider pool of movements – as a direct result of their varied client-base – to provide a more optimized set of freight movements.
Moreover, Shippers are increasingly adding another layer. They use an (internal or external) “Control Tower” approach, where a 4th Party Logistics provider (4PL) or Lead Logistics Partner (LLP) provides them with an overview of the operations (on top of the various 3PL activities) that improve visibility, merge activities over multiple regions, and hopefully optimize even further the logistics process.
With recent Artificial Intelligence (AI) and Machine Learning (ML) technology development, we can improve and modernize these operational layers and use data in order to offer smarter solutions. In particular, as the data points grow from Shipper, to 3PL, to 4PL activity levels, the capability of these technologies can make the model more and more efficient. (The “old fashioned” approach was to pre-determine routes, place orders onto those routes, and then send shipments on those allocated routes)
The main problem of the existing approach is that they are conceived based on the Shipper’s logic, but don’t focus enough on the end point; the Customer/Consumer side.
This is exactly what we want to change! We want to introduce a new design and implement a new paradigm by introducing some innovative mechanisms.
Apart from the “easy” FTL movements, there is a vast number of LTL. Groupage, and Parcel shipments that must be operated. Many point out, and rightly so, that the pandemic has accelerated the growth in Parcel particularly – following the growth of e-commerce and direct-to-consumer models.
But, however futuristic this might seem, can or should the paradigm be broken?
Let’s ask ourselves:
- Is there a way to re-approach logistics collaboration and render it more viable?
- Is there a way to include environmental concerns and benefits in this new approach of ours?
Needless to say, the suggested approach is entirely Customer / Consumer-oriented, and that all logistic methods & processes work from the Customer / Consumer demands backwards!
Let’s just imagine for a moment that a consumer goes online and places an order with two separate online retailers, whose parcel operations are handled by two different providers – one in-house and one outsourced.
What happens next is that both retailers use their own shipment optimization processes. We get TWO movements, at different times, to the same final destination, with each of those included in a “milk-run” of multi-stop deliveries…and often attempting up to a 100 deliveries per route, per day…
That means that TWO vehicles would move in a similar geography, going from initially 100% full to eventually 0% empty of goods. I.e., on average, we would have literally 50% utilisation. Such vehicles would literally criss-cross their movements, while covering two relatively similar and long routes.
But, what would happen if customers came first? What would happen if the final routing was based on the customer location and not on the source of the product?
Can we imagine a logistics scenery where cities, towns, or even larger regions in remote geographies, could create their own identified consolidation hub, so that goods are placed into those sites for their final mile delivery to customers? Just a SINGLE delivery and BOTH online orders shipped to the same final destination together, irrespective of the initial source of the product?
Assuming that the demand is still the same, we still end up with two delivery routes, but they would be designed wiser, and they would not overlap. The time needed to go from 100% full to 0% empty is halved, and the same vehicle can be utilised for other purposes or deliveries. The gains?
- Simpler route planning (as there are fewer final destinations)
- Reduced total delivery mileage covered
- A genuine environmental and fuel/cost benefit
With the power of Machine Learning or other Artificial Intelligence methods we can take in MULTIPLE sources of orders, broaden the carrier base and offer a truly collaborative last-mile delivery service.
If, on top of this, we add the requirement for more detailed final destination mapping (using something simple like What3Words) for all carriers, the opportunities and benefits would most certainly multiply!
It is time to place the environment in the centre of the conversation! It is time to re-calculate the cost-benefit equation from a Consumer / Customer perspective and to better design last-mile delivery methods.
Uptrend Labs is constantly working on these issues and paving the way in new directions. We are constantly researching for and applying innovative solutions and practical methods that respond to our client’s needs. Our current business and financial ecosystems are challenging, and their challenges we must face!
Please contact us to learn more.