Closing Last Mile Gaps: Addressing Process Industry Crisis in Volatile Markets
Process industry is facing unprecedented challenges in an increasingly volatile global market. From shifting global tariffs and changing trade alliances to mounting carbon reduction pressures and skilled workforce shortages, companies must now achieve both profitability and sustainability.
Many companies are seeking solutions in digital twins, but reality is far from simple. Why are digital twin projects that have received billions of dollars in investment failing to deliver the expected results?
David Arbeitel: A 40-Year Expert in Process Industrial Solutions
One figure provides clear answers to this question: David Arbeitel. He is a SIMACRO Board Member and former Senior Vice President of Products at Aspen Technology, brings 40 years of industrial solutions experience to analyzing why digital twin projects fail.
As the former Senior Vice President at Aspen Technology, he introduced ‘Industrial AI’ and engineering hybrid modeling to the global process industries several years ago, resulting in significant profitability and sustainability gains for Aspen Technology customers. In addition, he has accumulated deep expertise through various leadership experiences at global companies such as OpenText Vice President, ION Networks CTO, and Morgan Stanley Vice President, as well as experience in the AspenTech solutions field.
As hybrid models moved into production for use with online digital twins, Arbeitel found that many implementations failed to be adopted due to lack of scalability, usability, and convenience. In his presentation, “Closing Last Mile Gaps: Advanced Online Digital Twins” at a recent Simacro event, he identified fundamental disconnects between digital twin technology and operational reality based on decades of real-world project insights.
The 'Last Mile Gap' Problem
Arbeitel identifies the core issue as the 'Last Mile Gap', disconnections between ‘Design-Plan-Schedule-Operate-Report’ stages in process operations.
"The mismatch between plan and execution is causing significant yield losses. One refiner estimated $0.75~$1.00 lost per barrel due to last mile inefficiencies in a $70/barrel margin environment."
–David Arbeitel, a SIMACRO Board Member-
What is more serious is that this inefficiency has a negative impact on the environment, beyond simply losing revenue. One leading chemical company reported that greenhouse gas emissions increased by up to 5% due to inconsistencies between scheduling and operations.
Global Enterprise Cases: From Failure to Success
Many online digital twins lose their reliability within 6 to 12 months after deployment. Arbeitel refers to this as the “digital twin drift” phenomenon and identifies its root cause as the “failure to adapt” of digital twins. Most digital twins are optimized only for the conditions at the time of initial deployment and are unable to adapt to changing operational environments over time.
Arbeitel explains this through specific examples from leading global companies.
① ExxonMobil: The Optimization Trap
ExxonMobil's case demonstrates what Arbeitel identifies as the typical ‘optimization trap.’ The company struggled to maintain margin improvements due to differences between optimized plans and actual operational decisions. An ExxonMobil refinery located in the Gulf Coast region lost $22 million in potential profits annually due to frequent disruptions in its planning schedule.
Arbeitel diagnoses the core of this problem as "static optimization vs. dynamic reality" collision. From his perspective, no matter how sophisticated the optimization model, it will ultimately fail if it cannot adapt to real-time changing operational conditions.
② BASF: From Complexity to Opportunity
BASF's case illustrates what Arbeitel emphasizes as the ‘Unmodeled Transitions’ problem. Complexity in feedstock selection and campaign scheduling created off-spec and reprocessing cycles.
Inefficient batch switching and unmodeled transition processes resulted in 3-6% energy and material waste, meaning over 12,000 metric tons of additional CO₂ emissions per plant annually.
However, the BASF case is noteworthy as a successful improvement example. The next-generation solution, which maintains model reliability even during production, reduced rework rates by 40%, and one pilot project is expected to achieve margin improvements of $18 million and an 8% reduction in carbon footprint over 24 months.
Arbeitel's Revolutionary Solution: AI and Digital Twin Convergence
Arbeitel emphasizes that advanced online digital twins must leverage Gen-AI to close ‘last mile gaps’, creating a fundamentally new operational paradigm. He envisions Gen-AI and digital twins together forming a resilience engine that creates a continuous learning-operating loop.
Five core elements:
- Generative AI:
Enables predictive foresight and scenario simulation for plant decisions - Agentic AI:
Automates responses and guides operators through high-value interventions - Skills Gap Multiplier:
AI extends the capacity of lean, less experienced workforce - Integrated Intelligence:
Combines with real-time digital twins for autonomous optimization - Resilience Engine:
Together, AI and digital twins form a continuous learning-operating loop
Expected Revolutionary Outcomes from Last Mile Gap Resolution
When last mile gaps are resolved, the following improvements will be technically feasible based on verified pilot projects and case studies:
These improvements occur simultaneously through synergy effects rather than sequentially. Arbeitel emphasizes that addressing last mile misalignment can unlock millions in annual savings and reduce CO₂ intensity by 10% or more, representing enhanced competitiveness for the entire process industry.
Conclusion: The Beginning of Next-Generation Intelligent Plant Operation
Arbeitel's vision represents not simple technological advancement, but innovation that transforms the operational paradigm of the entire process industry. His insights demonstrate that for digital twin technology to truly succeed, three elements are essential:
- Technical Excellence: Real-time data integration and adaptive modeling capabilities
- Continuous Connection with Operational Reality: Intuitive interfaces that field operators can actually use
- Endless Learning and Adaptation: Systems that continuously evolve to match changing environments
These three essential elements are fully realized in SIMACRO's ProcessMetaverse (PMv). PMv leverages Gen-AI and advanced online digital twin technology to close the last mile gap between "plan and actual," enabling process industries to achieve profitability and sustainability in uncertain markets.
As Arbeitel's 40 years of experience demonstrates, technology's true value lies not in implementation but in continuous value creation and this is the essence of next-generation ‘Intelligent Plant Operation’.
With headquarters in Boston and Seoul, SIMACRO has completed over 90 commercial modeling projects across 40 companies since 2018. Collaborating with global technology leaders such as AspenTech, Emerson, and OLI, SIMACRO is committed to advancing digital innovation in the process industry.
About SIMACRODesignerDavid Arbeitel is a SIMACRO Board Member and former Senior Vice President at Aspen Technology, with 40 years of expertise in industrial solutions. He introduced Industrial AI and hybrid modeling to global process industries at AspenTech, and leverages his leadership experience at OpenText, ION Networks, and Morgan Stanley to provide digital twin technology strategies.
About David ArbeitelAbout David Arbeitel