Sensor battles senseless houseplant death

The San Francisco-based start-up has developed a 10-inch, stake-shaped sensor that obtains information about light, moisture, soil composition and other factors that can affect plant growth and health. The sensor is placed a few inches into the soil and connected to a computer, via a USB interface, where it downloads the information to PlantSense's Web site.

PlantSense can then tell users what they're doing wrong (too much sunlight, not enough fertilizer, etc.) as well as provide recommendations on what plants might grow best in a particular microclimate in a home or garden. Subscribers to the site can also keep records of the health histories of various plants and microclimates in their house and yard on the site.

"What we have developed is a plant lifecycle development platform," he said, co-opting some of the buzzwords he likely used while working at Cisco Systems and Xircom.

Killing plants is one of America's favorite pastimes. In 2003, Americans spent $18 billion on indoor and outdoor plants, not including grass. That's $160 per household, on average.

Roughly 14 percent of plants die in the first few weeks after being bought, and another 18 percent die within five months. That 32 percent mortality rate partly explains why Americans also spent $23 billion on fertilizer and plant food in that same year.

A big part of the problem is that most Americans have moved off the farm and into cities and suburbs years ago. Plants are also not great communicators: If a dog or cat is hungry, they will let their owners know; plants are more subtle with their signals.

The PlantSense stake will likely sell for around $49.95 when it comes out in the fall of 2007. Buyers will be able to access the PlantSense Web site for a year. A second year's subscription will cost about $20. Buyers only need to buy a single stake to test all of their soil. The battery in the stake lasts about a year.

That price range is in line with what consumers would expect to pay, according to Glenn. He polled about 200 consumers about whether they would like a product like this--81 percent said yes, and the median purchase price was about $50. An informal poll (conducted by the author at a lunch event) reflected similar results.

Glenn came up with the idea in January 2006 while getting a haircut and staring at some dying plants in a window. He asked his hairdresser why they were dying. She said she didn't know, but that it always happened.

By April, a prototype had already been designed. Initially, the PlantSense stake only provided recommendations on what kind of plant might thrive in a particular microclimate in the home or yard. A University of California botanist, however, told him he should add plant diagnostics for sick plants. To the sensor, those two jobs are identical.

The technology for sensing moisture in the soil, he added, comes from the same people who designed software for soil composition sampling on the Mars rovers.

Right now, the company is trying to land venture funding. It has retained a retail expert who worked with iRobot on the Roomba to craft a marketing strategy. In the first phase, the device will be available from the company's site and specialty retailers.

Over time, PlantSense will try to get onto the shelves at big-box retailers. Ultimately, the company would like to get plant growers to put a code on their plants, so that a consumer could get a soil/light/moisture score on a particular plot of land in their yard and then buy a plant with a corresponding rating.

But won't large retailers like Home Depot recoil against such a device? They make money right now by selling new victims to homeowners, after all. It turns out the big-box retailers may actually like the idea, according to Glenn. According to retailers, if people fail at gardening, they stop doing it; if they succeed, they buy more plants.