Event Brief: Honeywell User Group Meeting 2023
Honeywell held its 46th global annual user group meeting (HUG) in Orlando, FL from 19-21 June 2023. The theme of this year’s event was optimizing transformation. Attendees benefited from general sessions, breakout events by industry, as well as workshops and training. Over one-thousand three hundred attendees, representing two-hundred thirty two companies, attended this year’s event.
Axendia’s Sr. Industry Analyst, Sandra K. Rodriguez was invited to present the impact of FDA’s digital transformation journey on the life sciences industry.
Honeywell made major announcements at the event. For example, Honeywell Process Solutions announced its new Digital Prime solution, a cloud-based digital twin for tracking, managing, and testing process control changes and system modifications.
Additional HUG highlights are below.
Improving Cybersecurity and Operational Resilience in Life Sciencec Manufacturing
Cybersecurity continues to be an escalating challenge for life science companies. Paul Griswold, Chief Product Officer, OT Cybersecurity at Honeywell discussed the challenges of operational technology (OT) cybersecurity and best practices to reduce cyber risk and drive operational excellence.
Honeywell’s 2022 Industrial Cybersecurity USB Threat Report concluded eighty-one percent of malware can cause a disruption to industrial control systems. “The average cost of a cyberattack is now over $4M and OT networks are increasingly being targeted by attackers,” said Griswold.
Organizations operate in ecosystems, not one individual system – i.e., R&D, quality, supplier quality management, manufacturing operations, post market surveillance. “If one of these systems were to go down, they will all be affected on the other side of the process,” cautioned Griswold while offering the following examples of recent pharma cyber-attacks, citing media reports:
- April 2023 – Ransomware attack on Enzo Biochem stole the personal data of nearly 2.5 million individuals.
- April 2023 – Cyber-attack on Evotec took systems offline.
- April 2023 – Sun Pharmaceuticals hit by ransomware attack that compromised some of its file systems.
- June 2023 – Eisai Group Japan hit by a ransomware attack.
In this session, we learned the overall cost of lost production can be in the hundreds of millions of dollars, depending on how long a system is down. Griswold pointed to five key challenges organizations face when it comes to OT cybersecurity:
- Insecure remote access, i.e., lack of third-party access control or multiple remote connections.
- Limited centralized visibility i.e., no visibility of OT threats and vulnerabilities and lack of an OT specific response strategy.
- Lack of situational awareness, i.e., lack of risk awareness and incomplete OT visibility.
- Shortage of OT cyber experts.
- Site level challenges, i.e., limited understanding of site-specific cyber risks, unapproved software/processes, and lack of controls for removable media/USB usage.
Griswold explained that while most organizations have security information and event management systems, many companies do not have this technology for OT systems. “We have to find people who know not only cybersecurity but also OT and can implement that in an effective way and keep in mind, OT needs are unique,” said Griswold. Honeywell’s Secure Media Exchange (SMX) reduces cyber security risk and limits operational disruptions by monitoring, protecting, and logging use of removable media throughout industrial facilities.
“OT needs are unique, and availability is the most important thing in OT. It [OT] is nothing more than IT plus physics… but the end result is a physical change of a physical device on the manufacturing floor,” continued Griswold.
Griswold recommends companies start with an OT cybersecurity assessment, asset and inventory discovery, and then process control network mapping. Know what is on your process control network – there are often rogue devices as well. “Map network connections and traffic flow between assets. Utilize network visualization tools – make sure browsers that can connect to the internet aren’t accessing sites like Facebook and YouTube throughout the day,” cautioned Griswold.
In his final remark, Griswold offered this advice, “OT cybersecurity is a lot like yoga – you’re never going to be done. It’s always going to require continuous improvement.”
Use Case – Saving Time with Modularization
Josef Herzog, Associate Director at the ZETA Group, presented a use case on modularizing deliverables. ZETA Group specializes in the design, manufacturing, automation, digitization, and qualification of customized biopharmaceutical facilities for aseptic process solutions, focusing on the modularization of life science projects.
“Modularization is key for faster project delivery, not only by using standardized software modules, but also by modularizing deliverables,” explained Herzog. Through modularization, Herzog reviewed how ZETA was able to refurbish existing buildings for a client in Austria while implementing technical solutions simultaneously at a greenfield site for the same company in the US during the pandemic. This project had an ultra-fast and extremely ambitious timeline (two years) and required a high degree of parallelization in engineering, procurement, and execution.” Noteworthy is that ten months into the project, the team was faced with the challenge of meeting the timeline during COVID lockdowns.
“One to one tech transfer from an existing plant in the EU to a greenfield site in the US, was not feasible due to outdated technology at one site and competing regulatory requirements at both,” explained Herzog. Module type package (MTP) standards supports plug and produce manufacturing. “MTP is standardizing modularization. MTP and MES integrations will further enable plug and produce manufacturing in life sciences,” said Herzog.
What Can We Do For You? Surprise – It’s All About Generative AI
The final session of the event offered Honeywell’s TrackWise Digital clients a chance to participate in a product innovation workshop. The goal was to hear from their life science users how the solution can support future use cases. To no surprise, generative AI was at the top of mind.
I participated in the generative AI working group which included (among others) Monica Hunter, Global Quality Director – Quality Systems Management and Bhavik Padmani, Quality Systems Manager at Alkermes; Rex VanHorn, Associate Director, AI at Boehringer-Ingelheim; and, Nicholas Morales, Quality Manager at Altria. Discussions were led by Dan Burns, Director of Product Management and Darrius Kilpatrick, Product Manager for Trackwise Digital. A total of nineteen companies participated in the innovation workshop.
Below are the key takeaways from the generative AI working group.
- No matter how well a human does a search and sets parameters, they don’t always search correctly. AI could eliminate this challenge specifically with recurrence reporting and deviations.
- Leveraging AI so that the system checks the human, and in turn the human checks the system.
- Avoid searching and approving “bad data.”
- Nomenclature – sometimes words are hyphenated, sometimes they are not. AI can be trained to recognize both. Product names from call centers are sometimes misspelled. Train the AI to identify potential misspellings.
- People are historically bad at writing investigations and conducting investigations. Understanding the output of an investigation as a regulator or as it is happening is different. Leverage regulatory frameworks and ensure investigations are written in a uniform manner to satisfy regulatory requirements.
- There are a multitude of acronyms in any organization – large language models need to be trained on acronyms and their meanings.
- AI can be leveraged for vendor management review. Vendors can quickly be classified as good/bad and risk-tiering based on historical data. Scorecards can be automated.
- Use AI for pattern matching. For example, suppliers who provide raw materials around the world or vendors who have received 483s.
- Track the AI to see how it is performing to avoid model collapse. Is it correct 95% of the time? Is it being overridden most of the time? The AI may need to be retrained.
- Rely on the vendor (Honeywell/Sparta) to make recommendations when the AI needs retraining.
- How will FDA be using AI and what will the expectations be from regulators in the future?
- There are potential challenges with unlocking dark data such as getting insights on issues you’re not staffed to handle – what if 1,000 CAPAs get autonomously initiated and you can’t respond to them?
- Prioritizing risk. AI will allow organizations to take a risk-based approach to complaints. Pre-analysis will show which complaints needed to be acted on immediately or are most likely to get escalated to recalls.
- AI insights, currently available from Sparta, can look for clusters retrospectively, such as records that are similar, even as far back as 15 years or however old the data is.
- Humans who understand the manufacturing process will still be required to understand why there’s product variability, for example – QA doesn’t necessarily know.
Live polling at the workshop revealed the top QMS paint points, in order, that all organizations who participated in the working group are facing:
- Quality System Management Review
- Continuous Improvement
- Inspection and Audit Management
- Real Time Release
- Product Disposition
- Workforce Knowledge Excellence
- Design Control
- Product Recall
In Brief
This was my first time covering the Honeywell User Group meeting. Life science company representation was heavily tied to the company’s acquisition of Sparta Systems at the end of 2020.
There are a number of initiatives underway at the FDA that will impact life science companies in the future in the way of additional regulatory reporting. For example, as part of its $7.2B budget request for fiscal year 2024, the agency is asking for a $11.6 million increase toward improving medical device supply chain and shortage programs. FDA is also proposing new legislation as part of the budget request and is seeking the requirement for manufacture notifications at all times (not just in the event of a public health emergency), as well as the authority to require and review risk management plans to help ensure that manufacturers are prepared for situations where their ability to manufacture product may be disrupted or may be insufficient to meet demand.
Axendia’s research shows 73% of companies say digital transformation initiatives within their organizations were accelerated as a result of the pandemic. The Agency has a leg up on the industry it regulates considering it set out on its digital transformation journey fourteen years ago (although the phrase at the time was ‘eTransformation.’) Our message is clear: don’t let FDA win the digital race in life sciences.
A special thank you to Dave Medina, chief marketing officer at Sparta Systems, a Honeywell Company for the invitation to attend and present at HUG23.
Contact research@axendia.com to schedule an Analyst Inquiry on this topic.
The opinions and analysis expressed in this post reflect the judgment of Axendia at the time of publication and are subject to change without notice. Information contained in this post is current as of publication date. Information cited is not warranted by Axendia but has been obtained through a valid research methodology. This post is not intended to endorse any company or product and should not be attributed as such.