IBM Releases Pretrained Watson AI Tools in Several Industries
In a significant expansion of the IBM Watson cognitive computing platform, IBM has launched "pretrained" artificial intelligence (AI) tools for a slew of industries including advertising, agriculture, automotive, building management, customer service, human resources (HR), manufacturing, marketing, and supply chain.
"The focus is on how AI can make each professional—across industries—more effective and more efficient," Kareem Yusuf, Ph.D, General Manager of IBM Watson IoT, told PCMag.
Watson is IBM's series of AI services and applications. By releasing this series of pretrained tools, Yusuf said IBM aims to help companies change the way they work.
"We decided to release this largest-ever AI toolset pretrained for industries and professions to help businesses re-imagine how they work," Yusuf said. "A key business advantage lies in tapping into organizational insights, historical customer data, internal reporting, past transactions, and client interactions. These elements are too often underutilized."
Offering pretrained solutions for various industries is a big deal, explained Rob Enderle, Principal Analyst at tech analyst firm The Enderle Group. "It represents a significant maturing of the Watson platform," Enderle told PCMag.
Sometimes companies that deploy AI hit a snag during the training period. Since Watson completes its training before companies deploy the technology, companies can execute a more efficient deployment.
"Training is where AI deployments get hung up," Enderle said. "Much of the initial work with developed AI is to create this training, which then, through machine learning, can be passed on to new systems, significantly lowering the deployment cost and time to value. This is a critical phase to maturing the platform and getting it closer to its operational and sales potential."
With the heavy lifting completed during the training period, Watson is ready to start producing targeted, industry-specific insights right away.
"Getting the system to this phase is anything but trivial. Once there, machine learning can allow the replication of an unlimited number of systems," Enderle said.
Here are six industries in which IBM Watson is now pretrained:
1. Agriculture
IBM released the Watson Decision Platform for Agriculture to enable farmers to gather data on weather, Internet of Things (IoT)-enabled tractors and irrigators, and satellite imagery to let companies generate predictive data on farms. AI-powered visual recognition capabilities let growers decide where to spray pesticides, determine the severity of damage from pests and diseases, and forecast water usage. Farmers also gain insights from temperature and moisture levels, as well as crop distress.
2. Human Resources
With recruiters looking to browse through resumes quicker than ever, IBM is now offering AI functionality for HR professionals. IBM Watson Talent lets recruiters analyze the backgrounds of top-performing employees to find candidates for new positions. In fact, AI could help reduce bias in hiring decisions, according to IBM. Psychologists helped IBM produce an AI scoring system, which lets recruiters quickly sort through candidates. IBM says it has used AI to refocus recruitment time for companies such as BuzzFeed and H&R Block.
3. Marketing and Advertising
IBM Watson Assistant for Marketing is a component of Watson Campaign Automation SaaS. The assistant lets companies evaluate their marketing campaigns, engage in more direct conversations with customers, and create a personalized customer experience.
Meanwhile, WEATHERfx Footfall with Watson allows advertising companies to design ads based on shifting weather patterns. AI, machine learning (ML), and cognitive computing in WEATHERfx Footfall with Watson allow triggers to continuously self-adjust based both weather conditions and brand needs. WEATHERfx Footfall with Watson incorporates data from IBM MetroPulse, a business intelligence (BI) app that provides insights from neighborhood demographics.
4. Manufacturing
The Watson toolset for the manufacturing industry will provide visual and acoustic inspection capabilities. AI technology will also allow manufacturers to predict when equipment failures might occur, as well as energy waste and product quality issues. AI will let manufacturers gain insights and deal with workforce attrition, skills gaps, and rising raw material costs.
5. Commercial Space
For the commercial property and real estate industry, IBM IoT Buildings Insights lets property owners and building managers use data to reduce energy costs. It also gives them weather data insights and historical property performance. IoT Building Insights connects data from IoT sensors, main meters, and submeters. By working with Watson AI, IBM IoT Buildings Insights lets property owners analyze occupancy patterns. Building Insights is an extension of IBM's Tririga Facilities Management portfolio, an app that standardizes real estate and operations data.
"By using AI, contextual models, IoT, and other sensor data, IBM IoT Building Insights consolidates, stores, and analyzes your data in real time, seamlessly improving building operations and giving you unique insights," Yusuf said.
6. Transportation
IBM also introduced Watson IoT Platform, a requirements management solution for Watson designed to help the automobile industry improve the quality of vehicles. The Watson IoT Platform connects sensors to improve lead time and productivity in vehicle manufacturing plants.
"As vehicles become more complex, engineering requirements are exploding," Yusuf said, noting the 100 million lines of code needed to build a car.
The data insights from the requirements management solution will help the automotive industry meet improved requirements, according to Yusuf. In fact, 47 percent of projects fail due to inaccurate or poorly written requirements, according to the Project Management Institute.
"We use the power of Watson AI to assess the quality of requirements prior to review and provide guidance on how to improve their quality," Yusuf said. "Watson uses natural language processing to analyze requirements text and suggest improvements leveraging best practices from industry standards. Getting requirements right the first time reduces the cost and cycle time of validation by senior engineers."
This article originally appeared on PCMag.com.