Company emphasizes programmable automation for repeatable results
For scientists, the lab is often a place of discovery, research and of course, long, long hours of testing. In fact, many labs struggle to find people to work the third shift, where overnight hours might require re-running samples or setting up for experiments in the morning. To take the burden off scientists, Columbus-based aBioBot is turning to machine learning and automation to guarantee accuracy and look for reproducible results.
“We are trying to automate the typical lab processes in biochemistry,” said Raghu Machiraju, co-founder. “If you look at most labs, they are reliant on manual labor. Research scientists have spent so much time locked away, running tests. So why not have general machines that can be programmed to do a variety of tasks to handle the mundane aspects of the lab so that researchers can do real science?”
At aBioBot, automation starts with a platform made up of three essential components. The first, Gan, which is Chinese for “senses,” is made up of cameras, sensors and machine vision and machine learning algorithms to allow scientists to keep an eye on what’s going on in the lab. For example, Gan can check to make sure a well plate is aligned during research and testing. The second, the bot, is where Gan is housed, and acts as the hand that performs the work. The final component is aBioBot’s LabBench interface, a software that collects data and records experiments.
“You can think of Gan as the brain, which is observing and controlling the machine,” said Jany Chan, protocol engineer. “Then the bot is like the skeleton, or hand that is doing the work, and the software is what ties everything together into a single package. Instead of buying many different lab machines that each performs a single task, you can just write a set of instructions. From there, we can get our machine to do just about anything, so you’re not locked into a single technique.”
The flexibility of aBioBot’s automation is meant to replicate a personal lab assistant, running tests related to DNA sequencing, amino precipitation and a range of biology and microbiology experiments.
“Automation is the way of the future,” said Machiraju. “It’s necessary for economic reasons in terms of productivity, but also for reproducibility — so you can run the same experiment and get the same results. I think the time is ripe for biology to be much more automated so that the lab can become an integrated place where platforms and data can speak to each other.”
With the software development help of Big Kitty Labs, Machiraju and his team are preparing to do just that by looking to use the technology in a variety of industries including pharmaceuticals.
“I came to Ohio because there was a lot of interest in what I was doing,” said Machiraju. “It’s such a friendly and welcoming environment. Folks are looking for companies that are capable of introducing cutting-edge technology that has promise. And they want to make us part of that puzzle.”