Locus’ awarded patent for machine learning models for predicting traffic

Software company Locus Technologies received a second patent this year that will benefit last-mile hauliers and their drivers, among other benefits.

Locus’ “Machine Learning Models for Predicting Time in Traffic” patent provides logistics providers with hyper-accurate estimated travel times and improved predictability in their last-mile deliveries by accounting for traffic patterns that were historically considered too dynamic to map.

The patent covers a unique technology that analyzes historical traffic data and predicts travel times between origin and destination. It also takes into account sub-variables such as day of the week and time of day.

Locus founder and CEO Nishith Rastogi said the data will come from a variety of sources, including internal, external, or a mix of both. Modeling can provide accurate predictions depending on availability, cost and accessibility, among other factors, he said.

“The larger the amount of data, the better the traffic forecasts,” said Rastogi. “Some companies operating on a large scale may prefer to do this modeling on top of their internal datasets, which they would have derived from their day-to-day operations and fleet management processes. However, others may depend on partners like us to solve this problem.

Also Read :  AI Hardware Accelerators For Machine Learning And Deep Learning | How To Choose One

“Take Locus, for example, one of the features offered as part of our dispatch management platform, route planning. Our data comes from the deliveries made, as we have GPS tracks of the drivers,” he added. “Through our patented technology and a variety of other dynamic routing algorithms on this data set, we are able to accurately predict the time between point A and point B by considering more than 180 trades and about 250+ hard and soft real-world constraints. This allows us to create the most efficient routes for all orders for our customers.”

Accurately predicting time in traffic is largely a feature provided by map providers such as Google. Such tech giants are in control of the Android ecosystem and can do so to an unimaginable extent.

According to Rastogi, Locus’ patented technology gives the company the competitive advantage of accurately predicting time in traffic with limited data available.

Also Read :  Are Humans or AI Better at Detecting Deepfakes Videos?

Locus Founder and Chief Technology Officer Geet Garg said in a press release that Locus’ goal with its machine learning technology is to enable companies across all industries to drive high-precision logistics operations in any region, reduce costs and improve business to achieve success.

Rastogi tells CCJ that all stakeholders across the last mile ecosystem, from companies to delivery partners or drivers, would benefit from this new patented technology.

“Better traffic forecasts would lead to more on-time deliveries, which would improve the customer experience and in turn lead to revenue growth for businesses,” Rastogi said. “Also, this will strengthen the workforce, ie the drivers. Accurate traffic volume estimation helps them create optimized routes and help drivers complete deliveries faster. This will increase driver productivity, morale and satisfaction,” ultimately benefiting the last-mile hauliers they work for.

And Locus doesn’t stop at better traffic predictions.

Also Read :  Deepmind’s Protein Folders Awarded $3 Million

The company launched an initiative in 2018 that provides regular training, end-to-end IPR filing support and more to encourage all employees to file patents on their own behalf. The patent “Machine Learning Models for Predicting Time in Traffic” is a result of this program.

This is the fourth patent granted to Locus in four years, and more are on the way.

Rastogi said the company is working on several projects with a healthy pipeline from a technology standpoint.

One of those things, he said, is improving the data quality that fuels the prediction algorithm to continue providing logistics providers with hyper-accurate estimated travel times to enable more on-time deliveries.

“We’re also constantly finding ways to improve our dynamic algorithm that powers our dispatch management solution,” he said. “The goal is to improve the quality of the data fed into our systems so that we are able to increase efficiencies in the field and enable growth across all fulfillment channels.”


Leave a Reply

Your email address will not be published.