Keep calm and let the TMS decide

Moving freight is a complex business, and variables like rates, reliability, and capacity can change with the weather, the season, or the latest retail craze. In an effort to get the most from their freight transportation dollar, many companies turn to transportation management systems (TMS).

A good TMS tracks dozens of key performance indicators (KPIs) so users can weigh the variables and pick the optimal carrier, rate, and route. But what if a TMS could leverage the power of big data and ultra-fast processors to remove humans from the decision-making loop? Such a system could analyze far more variables than any human could handle, refresh its records with real-time data, quickly calculate the optimal shipping method, and even act on its findings.

That vision is quickly becoming a reality, thanks to the power of computer analytics, experts say. Adding embedded analytics to a TMS platform allows shippers, brokers, and carriers to make decisions based on the data they're actually collecting, not just on the trends they think they see, said Monica Wooden, chief executive officer (CEO) and co-founder of MercuryGate International Inc., a TMS provider based in Cary, N.C. "We're seeing this really evolve," Wooden said. "More and more every year, it's getting more robust and real time. And that allows everybody to benefit."

As is so often the case today, the rising interest in advanced analytics has a lot to do with the e-commerce explosion. Retailers face mounting pressure to meet escalating demands for next-day delivery and omnichannel fulfillment, both of which carry significant costs, Wooden said. In response, logistics executives and chief information officers are pushing for greater use of data-driven technologies like business intelligence and data analytics to help trim time and cost from their supply chains.

The fast growth of sophisticated inventory-tracking networks has given them the reams of raw data necessary to achieve that objective. By pulling data from smartphone apps, global positioning systems (GPSs), and electronic logging devices (ELDs), supply chain practitioners can quickly determine a shipment's precise location and its delivery status.

But the possibilities go well beyond tracking. "It's not just improved productivity, but true decision-making," Wooden said. "With embedded analytics, you can take empty miles out of the supply chain, work with people in certain lanes, make sure containers are full, and generally help the world be a better place."

For example, embedded analytics could help a TMS automatically book space on a preferred carrier in the Atlanta-Tampa (Fla.) lane, then revert to a second choice if the first carrier doesn't have the needed capacity, she said. Or it could suggest efficiency enhancements—such as showing that a carrier would save money by making multiple stops along its delivery route, instead of scheduling multiple trips with partially filled trucks.

That's not to say that only automated systems can make these determinations. People working in manual transportation operations make similar kinds of judgments all the time. The benefit to using a TMS to handle basic decisions is that it frees up human specialists for more nuanced decision-making, according to Wooden. An automated TMS would not replace human employees, but enable them to concentrate on more advanced tasks, she said.

Wooden is not alone in her assessment. Adding embedded analytics or "machine learning" capabilities to logistics software will reinforce, not replace, the supply chain workforce, agrees Eric Gilmore, CEO of Turvo, a collaborative logistics platform provider.

"The value of machine learning is to augment human intelligence and make people super-human," Gilmore said. He cautioned, however, that this requires a certain amount of database maintenance and upkeep on the user's part. Adding artificial intelligence to a TMS will not produce decent results unless the software includes accurate, recent data, he warned. Most businesses keep databases full of unstructured information, which include duplicate entries that can cause database chaos.

"You need good 'data hygiene'," Gilmore said. "You really have to feel that data is strategic to your business, and you need data scientists to cleanse it. You can't even talk about making a machine smart if you don't do that first. It's like the old saying: 'Garbage in, garbage out.'"

Companies are now starting to realize that they can't manage warehouses full of inventory without hiring data scientists to manage databases full of information, according to Jim Vrtis, chief technology officer of New Plymouth, Idaho-based trucking loadboard provider Truckstop.com.

"Data is the fuel for a good algorithm, which drives machine learning," Vrtis said. "We're past the time when it was just important to store the data in a database. We now have to understand it and leverage that information to make better decisions."

That's where data specialists can help. "A good data scientist can draw conclusions from the data that are impactful and actionable," said Vrtis. "It's almost like the gold rush. People say, 'I have a lot of data; now I need to hire a data scientist to come analyze it, so I can find the gold and make money.'"

The best TMS platforms allow users to be creative and flexible in making better decisions and saving money, said Mitch Weseley, CEO of Shelton, Conn.-based TMS provider 3Gtms.

That need is particularly important in light of changes in the TMS customer base, Weseley said. Twenty years ago, big shippers dominated the market, accounting for the majority of TMS sales. Today, however, most of the demand comes from small and mid-sized shippers and third-party logistics service providers (3PLs), he said.

"Creativity is so important. Both shippers and 3PLs have more levers they can pull nowadays," Weseley said. "You can't look at all the options and manually figure it out. So a TMS frees people up to do the things that can't be automated."

With tools like improved algorithms, robust database-building capabilities, and embedded analytics, software providers can help TMS users reach new levels of creativity, industry experts said.

"Those things empower today's [practitioner] to handle more freight, be more efficient, be more productive, and grow the business," Truckstop.com's Vrtis said. "They can spend less time connecting the dots and begin to take a tactical approach to freight matching and to improving service levels. I think it's going to be really fun to see."

Powered by embedded analytics, technology could soon help solve many of the problems that vex the logistics industry today. "This journey is at Day Zero in terms of what's possible in building intelligent software that makes the human smarter," Turvo's Gilmore said. "And supply chain is the most fascinating application for these techniques."