While IEEE Member Ryan Gariepy was pursuing a bachelor’s degree in mechatronics engineering at the University of Waterloo, in Ontario, Canada, he and three of his classmates entered a robotics competition—which led them to launch their own venture. The four students, each possessing his own set of engineering and business skills, founded Clearpath Robotics in 2009 in nearby Kitchener. The privately held company’s revenues are now in the millions.
Clearpath, which started out by customizing outdoor vehicles for academic researchers, branched out last year to make self-driving vehicles for industrial warehouses. Between its research and industrial divisions, the company offers nine types of robotic vehicles including a catamaran and two aerial models.
Clearpath formed a strategic partnership with General Electric Ventures in 2013, and last year the company secured US $11.2 million in venture capital. The money is being used for product development and manufacturing.
Marc Tarpenning, a cofounder of Tesla Motors, joined Clearpath’s advisory board in March. The robotics company now has more than 500 customers in 40-plus countries, including General Electric and John Deere. It serves e-commerce, mining, and other industries, as well as the military.
Initial funding from angel investors helped the startup get off the ground. Clearpath used the money to manufacture parts and store inventory. The company broke even after 18 months.
“By the time we started raising capital for our self-driving vehicles, we were five years into a successful business,” says Gariepy, the company’s CTO. “Investors were looking to get into robotics, and because of our proven track record, Clearpath was an intriguing opportunity with relatively low risk.”
Artificial intelligence has been incorporated into the vehicles to make them more efficient. For example, they can avoid obstacles and determine the fastest route to take without human guidance.
The self-driving approach shares the same underlying technology used in Google’s driverless car. SLAM (simultaneous localization and mapping) technology uses lidar sensors to scan the environment and create two-dimensional maps, then stores the maps so the vehicles know where they are and how to get to their destination.
Gariepy says he expects more AI capabilities to be added. Meanwhile, he says, what makes Clearpath stand out from its competitors is how easy its robotic vehicles are to deploy.
“The question for us right now,” he says, is “How do we get as many of our robots out in the world as possible?”
READY TO ROLL
The Otto line of self-driving vehicles is designed to move materials around a warehouse. The Otto 100 transporter can carry up to 100 kilograms; the Otto 1500 up to 1,500 kg. SLAM and other AI software allows the Otto to travel the fastest route to its destination while avoiding obstacles. Each vehicle has front and rear lidar sensors, which provide vision for the route and identify things in its path.
“A customer would start by driving Otto around the facility to give it a tour,” Gariepy says. “As it moves around, the vehicle makes a map. Then the Clearpath system is configured to set up the rules of the road for that facility by setting speed restrictions and guidelines on where Otto can and can’t go.”
Whether you have five Otto vehicles or a fleet of 50, the centrally controlled system needs to be set up with only one vehicle, which can then communicate with the others. Otto transporters can be configured with an accessory such as a conveyor or a lift to make it easier to reach and pack items. With accessories, different vehicles serve different purposes.
The vehicles enable warehouse operators to reduce the time it takes to get merchandise to the loading docks. They also lower labor costs.
On the research side of the business, Clearpath has a number of vehicles for rugged mining, military, and agricultural environments. The flagship product in this area is the battery-powered Husky UGV (unmanned ground vehicle), which can be customized to fit a variety of applications. It can carry 75 kg. The smaller Jackal, which was designed with the U.S. Army Research Laboratory, has a built-in GPS and onboard computer. The electric Grizzly RUV (robotics utility vehicle) can carry up to 600 kg.
One of the startup’s earliest research platforms was the Heron USV (unmanned surface vessel). Among the 28-kg catamaran’s uses are collecting water samples, monitoring harbor conditions, and measuring sludge in storm-water management ponds.
Clearpath, which started with four students, now has 130 employees. The team expects to add 100 hires this year across all departments including engineering, marketing, and sales. Last month, for example, the company was looking for an electrical engineer with experience in circuit design and debugging, as well as a robotics systems engineer with experience in industrial automation and software development. Job listings are available at the company’s website.
More than 6,000 people have applied for positions in the past year, Gariepy says. “Finding the right hire is one of the most challenging parts of my job,” he says, “because we want team members to be a strong fit in terms of skill level and culture.
“We have a good understanding of what our deficiencies are and our short- and long-term goals. We know what roles we need to fill today so that we can build the company we envision for the future.”
Gariepy’s advice for entrepreneurs is to get a lot of input before starting a venture—particularly from potential customers. “Don’t be afraid of talking to people about your idea,” he says. “They’re not going to steal it.”
The biggest challenge he faces at Clearpath, he says, is deciding what’s best for the company. “There’s always a question of risk versus reward,” he notes. “What is the thing I can work on now that can make the biggest difference for the company? It’s difficult to make such calls.”
This article is part of a startup series introduced this year featuring IEEE members who have launched their own ventures.
This article appears in the June 2016 print issue as “Blazing a Trail With Self-Driving Warehouse Vehicles.”
This article is part of our June 2016 special issue on artificial intelligence.