Sheila Jordan is a strong believer in generative artificial intelligence and has already rolled it out to all 95,000 Honeywell employees.
“There's not a single part of our company that isn't thinking about next-generation AI,” said Jordan, who has served as Honeywell's senior vice president and chief digital and technology officer since 2020 after years of leadership roles at Symantec, Cisco and The Walt Disney Co.
Jordan acknowledges that there is ongoing debate between those who have bought the generative AI hype and detractors who haven't yet bought into the technology, but he's such an enthusiastic supporter that Honeywell has already commercialized 16 use cases for generative AI. To generate excitement within the company, Honeywell is asking employees to submit ideas for how to use the technology, and they've received hundreds of suggestions.
But Jordan is also keen to remain fiscally responsible when considering when to leverage generative AI: She prioritizes the biggest, most impactful ideas first that can increase revenue, boost productivity or make employees happier at work.
Generative AI is pervasive throughout Honeywell in various forms, but some of the earliest use cases came from Honeywell's relationship with Microsoft, which it relies on for its Azure cloud computing services. Approximately 5,300 Honeywell employees have access to Microsoft 365 Copilot, and more than 4,500 software engineers use GitHub to write 90,000 lines of code each week. Engineers are drawn to the increased productivity, and Honeywell reports a 65% usage rate for GitHub.
Honeywell has also deployed tools like Moveworks' AI Copilot service to automate some of its IT helpdesk requests, resulting in an 80% reduction in incoming tickets as virtual assistants can now handle the majority of the workload, with only the most complex issues being handled by the human IT helpdesk.
Honeywell also introduced a generative AI virtual assistant called Red, which can quickly pull up data from the company's archives, including 350,000 pages of product manuals and more than 50,000 internal articles. Red responds in more than 100 languages and is available to any employee with a laptop.
According to Jordan, one of the biggest problems companies like Honeywell face today is drowning in large amounts of data that isn't necessarily organized properly. Generative AI can take that data, including unstructured data from videos or Word documents, and organize it to gain better insights that “provide a whole new perspective on what's going on inside your organization,” Jordan says.
Her efforts to recognize data stored across Honeywell predate the generative AI boom, but were timely given the importance of data to making generative AI work properly. Four years ago, Jordan began streamlining the number of software applications used at Honeywell. Once a particular vendor, like Salesforce or SAP, was selected, that solution would be used across the organization.
As a result, there are more than 1,000 software applications in use at Honeywell today, down from 4,500 when Jordan joined the company.
Honeywell now manages all of its important and critical data in Snowflake, including reservations, billing, inventory, HR, and engineering data. The data-driven approach enables Honeywell to make faster decisions and more accurate forecasts, which allows it to get more creative with dynamic pricing — a tactic that allows companies like Honeywell to charge higher prices by taking into account factors like regional preferences, seasonality, and spikes in product demand.
“You can't have an AI generation strategy without a data strategy,” Jordan says.
Besides working with Microsoft, Honeywell is also looking at generative AI tools from AWS and is in discussions with companies like Google. “The truth is, I think it’s not so much the technology but how you use it that’s important,” Jordan said, adding that he believes all the big players will offer good AI technology.
That said, Jordan doesn't want to see too many large language models in use across Honeywell, as this would make it difficult to adhere to the company's responsible AI governance and security principles. “Too much technology can become difficult to manage,” Jordan says.
And while a recent Honeywell survey found that only 17% of industrial AI decision makers fully implemented their original plans, Jordan says he's been in the tech industry long enough to remember when companies were uneasy about letting employees use their personal phones for work. Companies eventually solved that problem, and will solve it again with generative AI.
“Not everything is going to work,” Jordan said. “Find your best use case and stick with it, because it's not going to go away.”
John Kell
If you have any comments or suggestions for CIO Intelligence, please submit them here.
News Packet
27% of Fortune 500 companies see AI regulation as a risk. With the lack of comprehensive federal regulation on AI, bills remaining fragmented in U.S. state legislatures, and the recent passage of AI legislation in the European Union, an estimated 27% of Fortune 500 companies cited AI regulation as a risk in recent filings with the Securities and Exchange Commission. Top concerns include the cost of complying with AI regulation, the inability to predict what the regulation will look like, and the scale of penalties that companies could face if their use of AI falls foul of the regulation, according to a Wall Street Journal report. Still, despite hundreds of AI regulation bills pending in state legislatures, including nearly 30 in California alone, surveys consistently show companies are moving full steam ahead with AI initiatives.
IBM cuts jobs in China as Western retreat continues IBM is cutting more than 1,000 jobs in China, CNN reports, the latest sign that geopolitical tensions between Beijing and Washington are affecting the future of the technology industry in the world's second-largest economy. CNN notes that some companies, including Microsoft, have fired or reassigned employees over national security concerns. Market opportunities for enterprise IT in particular have worsened, in part due to capital flows being prioritized for state-owned enterprises and a shift in funding toward hybrid cloud and AI technologies.
Cerebras takes on Nvidia with new AI processor. Nvidia's huge market share in the AI chip market continues to attract rivals looking to chip away at its advantage. The latest challenger is Sunnyvale, California startup Cerebras Systems, which this week unveiled a new chip that it claims “beats” rivals at running AI models and generating responses (called “inference” in AI lingo). The chip will be offered as part of a computing system that companies can buy and run on-premise or access as a pay-as-you-go cloud service run by Cerebras. Speaking at an event in San Francisco this week, Cerebras founder and CEO Andrew Feldman predicted the new chip will take enough market share away from Nvidia that it will “make them mad.”
Adoption Curve
The industrial sector is using AI, but the justification for the spending is often questioned. Nearly two-thirds of AI leaders in a recent Honeywell survey cited increased efficiency and productivity as the most promising benefit of AI technology. Improved cybersecurity came in second, followed by improved decision-making through real-time data generation, according to a survey of 1,600 industrial executives worldwide conducted by Wakefield Research.
Lower ranked benefits of AI include greater work flexibility, greater job satisfaction, and more time for creative thinking.
But the survey also found that nearly half of respondents (48%) said they need to continually justify the resources needed to support AI implementation. That may be shortsighted. Jordan told Fortune that the manufacturing industry has yet to fully tap into the potential of Gen AI to improve how goods are planned, sourced, produced and shipped to customers. “Gen AI has the ability to take that data and look at the entire end-to-end system,” Jordan said.
Job Radar
employment:
– Arc Institute is hiring an IT Director based in Palo Alto, Calif. Posted salary range: $208,400 to $257,500 per year.
– Kidango is hiring a VP of IT based in Fremont, Calif. Posted salary: $230,400 per year.
Hiring:
– Adyen announced that Tom Adams will join the Dutch payments company as CTO, succeeding Alexander Massey, whose term ends at the end of 2024. Adams joins Adyen from Cash App, where he served as head of engineering for four years.
– Horizon Aircraft Corp. has appointed Tom Brassington as chief technical officer, the latest in a series of hires as the aerospace manufacturer looks to bolster its engineering team as it works to design hybrid-electric aircraft. Brassington joins Horizon from Lilium Corp., where he was head of systems design engineering.
– Accruent named Christy Yurecka as CTO and Brooke Huling as chief product officer, splitting the roles of chief product officer and technology officer after their predecessor, Richard Ruhrig, was promoted to president of the software company. As CTO, Yurecka will oversee product development and engineering operations.
– Careington International announced the appointment of Anthony Mustoe as CIO. Mustoe has 25 years of IT leadership experience in the healthcare, biotech and telecommunications industries, most recently serving as CIO at clinical diagnostics company SomaLogic.
