(EDITOR’S NOTE: This is another in a series of seven articles spotlighting the start-ups that comprise Cohort 3 of Oak Ridge National Laboratory’s “Innovation Crossroads” program.)
Jesse Claypoole is focused on developing a camera that does not exist in the market now so that smart machines can become even smarter. Whether it might be in autonomous vehicles or more mundane farming applications such as determining when tomatoes are ready to be picked, his technology holds the promise of impacting many business sectors.
The Richmond, VA native is another of the Fellows in Cohort 3 of Oak Ridge National Laboratory’s (ORNL) “Innovation Crossroads” (IC) program. After spending five years at the State University of New York Polytechnic Institute earning his Ph.D. in Nanoscale Science, Claypoole is very happy to be in more temperate climates while he works for the next two years on his novel technology.
“The camera that I’m developing will tell you three things” about an item, he says. Those three characteristics are what it is made of, where it is, and what kind of shape it has. How does it make those determinations? Claypoole says the answers come from assessing an item from both depth and spectral information without a reduction of the image quality.
“We’ll enable machines to see the what and the where in a single camera form factor,” he adds.
Claypoole, Founder and Chief Executive Officer of MantaPoole Technologies, has a provisional patent for the optics of the solution he is developing that he has named Plenoptics 3.0. That is the focus of his work over the two-year fellowship at ORNL.
In essence, you might say he is developing a tool that enables smart machines to become even smarter in making the right decisions. What does that require? The answer is access to all of the information they need and an assurance that it is correct. That’s no small task.
As these smart and autonomous machines become even more commonplace in the world, Claypoole says they will need an affordable sensor solution that can give them that needed information. Current approaches use standard 2D imaging sensors or expensive groups of sensors like lidar along with machine vision to interact with the real world.
“The problem with the current approaches is that they are either unaffordable to the average person or do not provide the correct amount and type of information for a smart machine to flexibly interact with the real world,” Claypoole explains. The MantaPoole Technologies solution is a new kind of imaging sensor that can capture multispectral and depth information in an affordable compact single camera form factor.
The business sectors that can be impacted are very diverse.
Claypoole’s first example during our interview was agriculture, specifically certain crops that can only be picked by hand. He said that broccoli is picked based on its shape, while strawberries are picked based on their color. The Plenoptics 3.0 since camera solution would allow farmers to know the precise times to pick.
Other potential areas could be food inspection, inspection of welds, and virtual reality where the technology would allow the use of just one camera rather than several in a headset.
Claypoole expects to spend the first 18 months of the IC Fellowship enhancing the camera before launching pilots to determine the best application areas. That period of time will also help him determine how much software MantaPoole Technologies needs to develop to bring its product to market.