By Linjing Cheng
It is commonly known that the rise of the computer has led to the end of many manufacturing and other jobs in America. But a handful of young software engineers started to see that cognitive non-routine jobs could also be automated in such fields as the communication part of businesses. Algorithms are the key to their inventions.
James Steinberg and Praful Mathur both graduated from Northeastern University a year ago. They met at a college Hackathon. Steinberg’s poster of Mark Zuckerberg amused Mathur and his club: “If you want to be Mark Zuckerberg, we can teach you how to be him.” Mathur and his team were trying to invent self-driving car services that move things around Boston. Steinberg, majoring in robotics and working on robots for warehouses, connected naturally with the group.
Couple of months before graduation, Steinberg visited Mathur’s self-driving car team in New Jersey. Then, on the day of his graduation, Mathur got a call. Steinberg said on the phone, “Hey man, I’m actually gonna move all my stuff into your house tomorrow.”
Mathur still found it funny today of how quickly that happened. He responded to Steinberg’s plans to move in with, “’OK, I’m gonna tell my parents,’” He added, “James is very decisive. He does some experiments and knows what to do. It was really helpful to have him on the team.”
With a background in robotics, they were discussing developing robots for warehousing and moving of manufactured goods. However, the people in warehousing and distribution kept thinking Mathur and Steinberg’s robots are like the ones in science-fiction movies.
Mathur said, “They thought that robots can solve every problem that they ever had, like labor problems, problems with their wives, ha, literally, you name it.” Mathur explained that robots could only solve a narrow set of problems that people would program them to do. They are good at doing “specific things repetitively.”
A friend’s connection with an entrepreneur who had built warehouses and factories offered advice that defined what later became their company, Derivatived. He said to Mathur, “What you are solving is a logistics problem. You are not solving anybody else’s problem here. Moving things around, to different people, at the right time and place, that’s the basics of logistics.”
The small group of staff then created the entire new online logistic service company, Derivative/d. Its website operates several functions to help production companies find warehouses, and that work is all done by algorithmic calculations.
Algorithms sort out patterns and predict outcomes related to the data set given. This attribution would change the game rules for warehousing, and make things convenient for the product companies. The warehouses, having been the rule makers in logistics, will go through changes.
An economist who was studying job computerization said in 2013 that algorithms, because of their ability in data pattern recognition, can readily substitute a wide range of non-routine cognitive tasks.
This study determined that logistics — the planning of locating and moving goods — is one the first occupations to soon be automated with algorithms. Mathur and Steinberg’s small company, Derivative/d, hasn’t made much noise in the business world as it is so far the only one of its kind.
At one point, the two software engineers were trying to sell robots that move things in warehouses. Soon, Mathur found that the warehouse owners imagined that the robots could do things cooler than what he designed.
It was the beginning of their many pushbacks from a non-technological industry. The date of putting robots into warehouses went from two to three years away into unsure. Meanwhile, the group was thinking about what they can do with their self-driving cars and where could this be most useful.
Derivative/d finds the right warehouses to meet production companies’ needs for short periods of time. Why can’t manufacturers search on their own like we search for our apartments? Surprisingly, according to Steinberg, some warehouses only offer long-term leases, and manufacturing companies may not know the total logistics cost until they sign a lease and ship the goods.
All third-party logistics providers follow up requests with phone calls. Lots of the sales people didn’t pick up the phone call.
Later on he added, “It helps especially startups that don’t have experience with logistics and their own supply chain.”
That was from the manufacturer to the warehouse. Things are more arbitrary from warehouse to the customer. “For shipping to the customer, the warehouse is going to find out how much is the actual price, and they will tag on this margin,” Steinberg said. “So they kind of just make it on the spot.”
To locate the best warehouse for a specific manufacturer, Steinberg and Mathur said they have to take into consideration the warehouse fee, the local labor fee, transportation fee, as well as changes in location of a company’s’ customers. On the warehouses’ end, multiple short leases replace one long-term lease. It works as well.
Data mining is essential to their work, but getting data has been quite difficult in this case. Not all warehouses have a fixed method for pricing products.
3G Warehouse Storage and Transportation Services, an average size storage company that owns four warehouses, is one of Derivative/d’s partner. Lauren Nichols, owner of the company said that to get a quote, a price rate, from her takes 24 hours and shorter. She said that the quotes include”picking up the goods at the warehouse, storage handling, order processing, materials like boxes and labels.” As for the shipping rate from the warehouse to the end customers, Nichols said, “sometimes it does, unless it’s shipping with UPS, USPS and Fedex.”
Steinberg and Mathur want their company to use software to innovate storing and shipping hardware. “The biggest boom right now is hardware. Other than that, you are all over the place. You have automotive parts, wheel chairs, food. It goes everywhere.”
It is a great time for algorithms as well. “The pace of technology today of advancing what society can do is happening amazingly fast,” Mathur said. “Distribution of innovation is instantaneous.”
Steinberg explained what peoples’ jobs are like working on the communication: “If you are a product company, you will have to call hundreds of warehouses to get one that makes sense. You will have to constantly contacting your warehouse, saying that your new product is coming in, asking what’s gonna happen… new orders coming in. Sometimes they have software systems to do that, but lots of times it’s manual back and forth, emailing and phone calls. You have to have a long term relationship with your warehouses, but now we are eliminating that.”
And that instantaneous innovation is happening in using algorithms for better and faster communication in the logistics industry, thanks to Mathur and Steinberg.
“Given their software, logistic people who work at big companies and startups can transit their jobs into improving distributing resources,” Mathur said.
Right now in Amazon’s warehouses robots are now moving the products. Humans come in when something goes wrong, “Is the future robots gonna have some impact on our business? Absolutely. It will make our service cheaper,” Mathur said. “That’s something in the future. Right now we are focusing on automating the communication.”