Listen to a group of “green thumbs” talk about growing tomatoes and you’ll quickly realize how competitive an endeavor this is. Each gardener has their own special tricks for coaxing the best tomatoes off their vines. They’ve taken years to uncover these secrets, watching countless plants grow or die based on varying conditions and techniques.
Such information is valuable, but human experience comes at a high cost, both in terms of the years and extended effort required to earn it. Even with experience, people are limited in their ability to observe thousands of plants with a high attention to detail. People also get tired and can make mistakes both in what they observe and the conclusions they arrive at.
Nature Fresh Farms, one of Canada’s largest independent greenhouse produce growers, is an industry leader in their work with cutting-edge technology by utilizing video and artificial intelligence (AI) to analyze its greenhouses and drive better business decisions.
Ideal greenhouse conditions are easier to maintain with constant observation of plants. Computers never tire of watching nothing happen.
One of the critical advantages with this approach is that AI systems are infinitely patient. Ideal greenhouse conditions are easier to maintain with constant observation of plants. Computers never tire of watching nothing happen. They are also capable of immediately catching small changes. For example, a person may fail to notice holes in leaves for a few days. A computer will be able to observe bugs eating the plant as it is occurring. This information can be acted upon quickly, bringing in robots to inspect the plant. AI can also distinguish different bug types, recommend good bugs to bring in to control an infestation, and assess when the bad bugs have been eliminated.
A smart organization like Nature Fresh Farms can build on its expertise for growing plants at a highly accelerated rate. With recorded video-based data, growers can review the lifetime of different plants to determine which factors most likely impacted their growth. AI takes what is called “intuition” in an expert grower and creates concrete rules and methods for achieving desired results. These systems learn at an exponential rate compared to humans. This information can also be shared among different deployments, further accelerating the learning process.
Advanced Senior Care
Many of the lessons learned in the greenhouse can carry over to other applications. Consider senior care, where computers with AI can be used to assist with our older population. One of the primary purposes of senior care is to help elders maintain their autonomy.
Frequently, seniors at risk are moved out of their homes into facilities where it is easier to watch them. For seniors who stay at home, someone needs to check in multiple times a day to ensure their safety. Today, systems have been developed comprising sensors and cameras to collect information about seniors living at home. With infinite patience, these systems watch seniors and learn their routines and behaviors. The more they learn, they more accurately they can automatically verify seniors are all right. An AI-based system can take over much of the burden of watching over seniors. For example, the system could verify that the senior got out of bed and made coffee. When an expected behavior is missed, the system can then perform further assessment, potentially leading to an alert for outside aid.
The extended freedom and autonomy this gives seniors is not the only benefit. Their quality of life is also improved. For example, instead of calling several times a day to make sure their parents are all right, children of seniors can reach out and share their lives. This is because they can be confident that care details are all under control and let their worries go.
There is also the benefit of community learning. Systems that observe seniors learn about individuals but also learn about elder care in general. These lessons are expensive for caregivers to learn as they require years of experience. With computers, we will be able to learn so much more about taking care of our parents with dignity and excellence.
A system watching thousands of subjects has a lot of “memories.” It isn’t always immediately clear which of these “memories” will be important later. As a consequence, learning based on video data requires a great deal of local storage that can scale-out to meet increasing capacity demands. In tandem with storage requirements is the need for accelerated compute options, such as those available from NVIDIA.
Dell Technologies is perfectly placed to guide OEMs through their AI journey with specialized services like developing an Artificial Intelligence strategy and a portfolio of industry-leading AI solutions including the Dell EMC Ready Solutions for AI and Reference Architectures for AI. These AI solutions combine flexibility and Dell Technologies’ expertise to deliver NVIDIA GPU accelerated compute complemented by Dell EMC Isilon’s high performance, high bandwidth scale-out storage solutions. The result is simplified data management for training the most complex deep learning models, no matter how large a “memory” your system needs to have.
Dell Technologies is at the forefront of AI innovation to help you make tomorrow possible. Learn more about how AI is changing how we care for ourselves and others from our interview with theCUBE TV at Dell Technologies World as well as from the on demand session on Platform considerations from Proof-of-Concepts to Large Scale AI Deployments.