October 1, 2017 (Vol. 37, No. 17)

Gail Dutton

Telos Scientific Suggests Putting the Integration of Laboratory Automation on AutoPilot

Telos Scientific is the developer of AutoPilot, a technology that can allow scientists who lack programming expertise to operate sophisticated automation systems. The AutoPilot interface can accept procedural information, much like a laboratory notebook, and then use artificial intelligence to interpret the input and autonomously control the available robotics, such as liquid handlers. Modules are available for tasks such as serial dilution, the preparation of samples for pharmacokinetic/pharmacodynamic analyses, the identification of hits to be picked, and genomic DNA normalization and aliquoting.

Purchasing laboratory automation equipment is only the first step in a lengthy, expensive process. The next step is integration—the incorporation of a vendor’s off-the-shelf product with an end-user’s real-world and often unique operations. Integration often requires end users to muster what is, for them, novel and specialized expertise before they can obtain reliable results.

End-users may find it easier to meet the automation integration challenge if they “go meta,” that is, if they take their automation activities to a higher level of abstraction. Specifically, end-users may want to try automating their automation implementation processes, not just their laboratory processes. This option is emphasized by Telos Scientific, the developer of AutoPilot laboratory automation software for low- to medium-throughput laboratory processes.

“AutoPilot’s proprietary artificial intelligence engine collects assay protocol information from scientists, analyzes the laboratory’s available instrumentation, and then programs those instruments for the optimal parameters required for those protocols,” explains Jesse Campbell, Telos Scientific’s founder and director of custom solutions.

Learning from Integration Failures

Before founding Telos Scientific in 2009, Campbell picked up a bachelor’s degree in microbiology and molecular genetics and embarked on a career in laboratory automation. Besides developing consulting and training expertise, he began facilitating company transitions from manual to high-throughput assay processing.

He learned firsthand that “once the laboratory automation equipment is in place, a significant amount of time, money, and expertise is required to produce any good scientific data.” He also came to realize that the laboratory integration challenge was all too common. At many companies, it is assumed that automation systems are as easy to use and as customizable as other laboratory tools.

“Laboratory automation systems are a form of robotics, so you need cross-disciplinary expertise in engineering, programming, and science,” Campbell says, adding that many scientists lack this combination of skills. When scientists attempt their own programming, they experience many delays and frustrations.” As a result, Campbell observes, “a high percentage of laboratory automation systems never go into production.”

Another challenge is that turnkey systems really aren’t turnkey. “Each laboratory,” Campbell points out, “does something a little bit differently.”

When he formed Telos, Campbell hoped to realize a vision: improve the integration and automation efficiency of the robotics and instrumentation adopted by small- and medium-throughput laboratories. More specifically, Campbell meant to improve the precision and accuracy of assay outcomes. To that end, Telos Scientific has developed novel technologies that aid in automation implementation.

Helping Scientists Avoid Programming

Correctly programming laboratory automation solutions traditionally requires cross-disciplinary expertise. Incorrect programming can result in increased time-to-production and inaccurate results.

Telos Scientifics’ AutoPilot relieves scientists of programming responsibilities, saving significant time. “Depending on the complexity of the protocol and the expertise in your laboratory, it’s not uncommon to spend months putting a protocol or groups of protocols into production,” asserts Campbell. “With AutoPilot, our clients typically achieve the same results within days.”

Assay development is performed directly in AutoPilot’s LabNotebook using a user-friendly, scientifically familiar format. Importantly, before starting a run, “AutoPilot’s artificial intelligence engine confirms that it knows how to perform that task,” explains Campbell. “If it doesn’t know how, it alerts the user and suggests alternative tasks. A lot of work went into ensuring it doesn’t try to do something it doesn’t know how to do.

“Our products and services are all geared toward learning what scientists need to achieve. We provide consulting services, software products, and novel, proprietary technology that eliminates the need to bring in cross-disciplinary expertise.”

Automating Validation

Telos Scientific introduced an automated validation module for AutoPilot last spring. By year’s end, the company plans to launch version two, which will include an improved user interface.

“There’s often a large disconnect among quality control specialists, scientists, and their programmers,” Campbell says. This disconnect complicates the validation process and may cause faulty results.

AutoPilot’s Automated Validation Module, as the name implies, automates the validation process. “AutoPilot dynamically creates all the test cases needed to test users’ automated protocols,” he declares. “It is also designed to perform the validation run and test the liquid handler under the exact conditions required for the scientific protocol, collecting data and creating the validation documentation.” These capabilities are, according to Campbell, unique in the market.

The module has already helped customers meet their validation challenges. “One client with multiple automated systems dedicated two scientists and four years to process validation, but hadn’t gotten its laboratory automations systems into production,” recalls Campbell. “AutoPilot validated the automations and put those systems into production within two weeks.” He adds that another client reported cutting process validation times from as long as six months to a mere two days. (Obtaining signatures on the paperwork took another two weeks.)

Delivering Scientific Results

Telos Scientific also offers more customized services. In 2016, in Seattle, WA, it opened an engineering facility to customize the hardware components needed by scientists. “They tell us what they’re attempting to automate—next-generation sequencing, for example—and we help them automate it,” Campbell says. “We’re also developing several additional software tools that use novel technology to shift the focus from getting laboratory automation to using laboratory automation as just another tool to achieve scientific results.”

Telos Scientific

Location: 3636 Camino del Rio N, Ste 102, San Diego, CA 92108

Phone: (858) 334-3630

Website: telosscientific.com

Principal: Jesse Campbell, Founder and Director of Custom Solutions

Number of Employees: 6

Focus: Telos Scientific provides laboratory automation software, hardware, applications, and expertise for low- to medium-throughput laboratory processes throughout the world.

 

Previous articleMegarafting Animals Rode from Japan to U.S. and Canada after the 2011 Tsunami
Next articleImmunotherapy Response Predicted by Liquid Biopsy