Workforce Research

Research in Workforce Management and Learning is focused on creating innovative learning technologies to enhance student performance ; representing and validating employees’ skills, to facilitate matching people to opportunities; and applying supply chain modeling to optimize human resource planning and management
In the last few years, changing business conditions have increased the pressure on our clients to optimize the management of their workforce. To remain competitive, they need to deploy people more flexibly and at lower cost. Global resourcing presents unprecedented challenges in the volume and speed with which they conduct hiring and training and invest in retaining their workers. Leveraging our successful history in applying supply-chain modeling to achieve significant cost reductions in IBM procurement, call center operations and Business Continuity Recovery, Research is applying supply-chain modeling and tools to characterize human capacity, forecast demand for human resources, and optimize the matching of people to jobs within IBM. These models and tools can be hardened into service assets to help our clients optimize the management of their workforce -- reduce labor cost, retain critical skills, and deploy people more efficiently.

Our work spans research in Workforce Management and Learning and is focused on:

• Creating innovative learning technologies to enhance student performance and increase learning through personalization and contextualization
• Representing and validating employees’ skills, to facilitate reskilling and matching people to opportunities
• Applying supply chain modeling and mathematical algorithms to optimize human resource planning and management

• Dynamic Learning Experience (DLE): This project addresses on-demand learning for employees, who need to learn topics just in time, while performing a work-related task. The system dynamically assembles courses out of a repository of learning objects, respecting the user preferences. The team has deployed the system within IBM and has conducted a pilot with the Defense Acquisition University of the US Department of Defense.
• MAGIC (Metadata Automatic Generation for Instructional Content): This project has developed tools to automatically generate metadata needed for SCORM learning objects, with the goal of substantially reducing the labor involved in the tagging of learning content for access, search and reuse. The project provides an environment to process learning content using IBM Research's leading-edge text and video analytics. This project was funded by a grant from the Department of Homeland Security.
• Literacy Tutor: This project is developing a web-based system that allows both children and adults to increase their literacy skills by practicing reading. The program, called Reading Companion, uses an interactive tutor character to present learners with e-books and speech recognition technologies to 'listen' to the reader practice and provide immediate feedback.
• Expert Tracker: This project is developing a system to support collaboration, knowledge reuse and question asking. Expert Tracker provides experts with fine-grained control over when they appear as available for collaboration and expertise sharing.

• Skill Affinities: The goal of this project is to address the problems encountered when trying to match available human resources to open positions. Currently, only exact matches are returned. The project will create a skill affinity computation and embed it in a search tool to allow for approximate searches.

• Resource Capacity Optimizer: The project has developed a tool to compute "gaps" and "gluts" in human resources, based on business logic and recommend staffing. The tool optimally matches supply against demand, indicating excess and shortages. It has been deployed as part of IBM's workforce management process. Other Research assets support resource capacity planning optimization under demand uncertainty.
• Workforce Evolution and Long-term Planning: The project is developing a tool that provides a long-term view of the evolution of the workforce to determine optimal business strategies, organizational structures and long-term operational policies. Given forecasts of demand, revenue, supply and cost, the tool allows users to determine the optimal workforce evolution over time in order to make strategic long-term planning decisions

• Engagement Profiling: The project creates standard staffing templates for engagements, that enable project managers to develop early staffing plans to improve planning. The project uses statistical clustering methods to group historical engagements that have common patterns in their skill mix to create a taxonomy for projects on the basis of their resource requirements. The templates are being deployed within IBM.
• Short-term Demand Forecasting: The project focuses on forecasting the high volume of demand in emerging countries, based on an analysis of historical load patterns and the application of a load factor to the amount of resources currently being requested.

• Individuals to Open Seats: This project has created a decision-support tool to optimally match moderate to large numbers of resources to open positions. The tool uses a constraint satisfaction paradigm to optimize the matches globally, rather than deal with them on first-in-first-served model.
• Individuals to Shifts: This project has developed an asset for the automatic creation of schedules for shift workers that minimizes payroll costs while balancing required service level on the one hand and compliance with labor regulations and rules on the other. The tool, called SWOPS, has been successfully applied to customer engagements, with resulting 10%-13% payroll reductions.

• End-to-End Workforce Management: The goal of this project is to develop an architectural framework, analytical methods, and optimization algorithms, for supporting an end-to-end workforce management process. The project includes the development of several advanced analytical capabilities to support processes throughout the workforce cycle, including demand forecasting, capacity planning, gap/glut analysis, skills/resource matching, available-to-promise, available-to-sell, and demand conditioning. These analytical capabilities are implemented within a highly scalable service-oriented architecture and seamless data management and integration, which enable flexible solution reusability. The first prototype of the system, called “OnTheMark”, is currently under deployment within an IBM business unit.

Aleksandra Mojsilovic; Amy G Katriel; Bonnie Ray; Daniel Connors; Donna L. Gresh; Heng (Harriet) Cao; Jianying Hu; Keith Grueneberg; Kevin Singley; Jennifer Lai; Rohit Lotlikar; Moninder Singh; Murray S. Campbell; Mark S Squillante; Mayank Sharma; Yehuda Naveh; Peter Fairweather; Yael Ravin; Amnon Ribak; Segev Wasserkrug; Akira Tajima; Ta-Hsin Li; Tracee Wolf; Hao Yu

Research labs involved

Almaden Research Center, Haifa Research Lab, India Research Lab, Tokyo Research Lab, Watson Research Center

Additional information

Demo (requires Java Runtime)

Download Java Runtime