When we had a car shipped to us from California a couple of years ago, I remember the trucker unloading the vehicle and then working his laptop, looking at leader lines and trying to determine if he could pick up a new load in Seattle so he wouldn’t have to drive back to Southern California without a load.
SEE: 5 Internet of Things (IoT) innovations (free Pdf) (TechRepublic)
It was a tedious process: He looked at several possibilities and made a series of phone calls. Eventually, he had a next-day load lined up, although pickup was over an hour north of us that he would first have to travel before he could turn around and make the trip back south.
I thought to myself then, that there must be an easier way to do this, and apparently that’s beginning to be the case. The breakthrough is not because of revisions in logistics methods, faster trucks, or more cargo. It’s being driven by technology advances that have to do with big data.
You can stream unstructured big data such as locations from GPS devices and Internet of Things (IoT) data from smartphones that can tell you the speed and location of specific vehicles. You can combine this with incoming data from shipping and logistics pickup and drop-off points. As the data gets processed, you can use algorithms and machine learning to calculate the optimal routes for trucks carrying specific types of cargo.
SEE: TechRepublic Premium editorial calendar: IT policies, checklists, toolkits, and research for download (TechRepublic Premium)
“What we’re using is relevance formulas for big data that have already been used in other applications,” said Ziad Ismail, chief product officer for Convoy, which provides a digital freight network. “We can rapidly process dynamic data that changes second by second and answer questions such as, ‘What is the best truck to match up with my load, if I’m a shipper?’ ‘What is the best route for me to take if I’m a transporter?’ ‘Are the distribution points I’m heading to ahead or behind schedule for unloading?’ ‘How can I maximize my trip by minimizing the empty miles I travel when I’m not carrying a load?'”
The technology depends on using the smartphones that drivers carry with them on their trucks. “You can think of these phones as IoT devices in their own right,” Ismail said. “These phones contain acceleration meters that can tell you how fast a truck is going, and GPS that tells you its location.”
Best of all, a carrier no longer has to wait to talk to someone before accepting a load. Instead, the system finds it for him or her. They can receive routes where delays are avoided—but if they do get caught in a delay to where a detention payment must be made, the system can do this for them through an auto payment function.
“That’s the advantage of the smartphone,” Ismail said. “So much sensor functionality is already built into it, and you can set up the phone with the cloud in less than five minutes if the driver is in good standing.”
SEE: 5 IoT lessons taken from the logistics industry (TechRepublic)
The Convoy technology consists of AWS cloud-based software, an assortment of data warehouses on the cloud from different vendors, a toolset for developing and tuning machine learning models, and a machine learning revision process that is executed every day.
“Especially since the COVID-19 crisis, supply chains change constantly,” Ismail said. “Because the supply chain is constantly changing, you must continuously revise your machine learning models to reflect the realities of what the supply chain looks like on any given day. What we’re seeing now is a new openness in shippers to use technologies that their internal systems are not able to support.”
That’s a major change of thinking for an industry where change historically has come slowly, and it contains a lesson for those who want to harvest big data and use the IoT that they already have under management: Do not overlook the IoT sensors and capabilities that are already built into your employees’ smartphones.