Building collaborative engagement
The Office of Business Partnerships at Purdue University’s Mitch Daniels School of Business creates and provides solutions to the business challenges of today and the future.
Dedicated Daniels School professionals streamline opportunities for industry partners and alumni to connect with students and faculty beyond the traditional classroom setting. The partnerships office helps companies solve business challenges through consulting-focused experiential learning projects across all business functions via course-based engagements or contracts with our centers.
Applying the power of data to serve industry needs is in the Office of Business Partnerships’ DNA. Moving beyond theory into real problem-solving is a call to action our office takes seriously. We understand organizations large and small face uncertainty, rapid technological changes and the continuous need to train employees and cultivate new talent.
Let’s get started, together.
TOP 8
Public University in the U.S.
QS World University, 2025
TOP 10
Most Innovative Schools in the U.S.
7 Years Running
U.S. News & World Report, 2025
BRANDS THAT MATTER
For the 3rd year in a row, Purdue is the only university on Fast Company’s Brands That Matter list.
There are so many ways we can work together to create and nurture a rewarding partnership that benefits your organization. Explore the opportunities that make up the core of a strategic partnership with Purdue and the Daniels School.
Enhance your organization's leadership capacity by investing in their development through a residential, online or hybrid degree program, certificate courses or a custom offering.
Collaborate with our centers of excellence and expert faculty researchers to gain access to top thought leadership.
Develop future team members while filling your most challenging positions. Your next hire could be that talented intern, a stellar full-time candidate or a transformational executive you meet through a tailored recruitment strategy designed for your company.
Join forces with our expert faculty and top students to create a consulting project that will define and test solutions for your pain points.
Drive industry and academic innovation through board membership, mentoring, guest lectures and curricular guidance.
The industry partner is a large welding equipment manufacturer with a low volume, high mix of machines. The current supply chain for welding machines has ≈ 13,000 unique parts across ≈ 800 different suppliers.
Today, the industry partner employs pull-system planning policy through our vertically integrated facility as a lean manufacturing strategy to reduce on-hand inventory. The company does not utilize traditional materials requirements planning (MRP) as generally seen in a push system.
The industry partner will share two different forecasting scenarios with our suppliers, which are referred to as the “Standard” and “Max” forecasts. The value provided is the forecasted monthly consumption.
Scenario #1 (Standard forecast) – A straight-line explosion of the sales forecast
Scenario #2 (Max forecast) – An algorithm based on the historical demand variability to capture a surge or peak monthly demand level over the trailing 12 months.
Quantifying energy grid load by geography is extremely challenging since the grid is more akin to an undirected graph network than a clearly defined set of start and end points. In other words, electrons are not tracked directly from the generator to the customer connection. This black box effect makes accurately forecasting grid behavior (such as swings in demand, load, surplus, etc.) challenging, particularly when it is desirable to forecast realistic behavior in a focused region (ex., by zip code).
The industry partner has approached this previously by leveraging an eXtreme Gradient Boosting algorithm (XGBoost) trained on energy market data for a region of interest to regress against outputs from PLEXOS (an energy grid simulation tool), using price as a surrogate for a demand metric. Applicable time scales for prediction: next minute, next half hour, next hour, next 24 hours, next week, next month, next year.
Proposed Methodology: Leverage open-source algorithms like (S)ARIMA(X), XGBoost, Meta’s Prophet, ETS, and relevant neural network architectures like RNN, LSTM, or Transformers to simulate and predict energy.
In this engagement, you will be helping to improve the company partner's predictive spend and pricing analytics capabilities. The idea would be that they use our historic pricing trends (not tied to volume) by category to forecast what spending would look like 12-24 months out. This will likely include the need to input cost data from publications like CDI or IHS to analyze how trends in input costs correlate to our actual costs over time. We can then use that information to proactively buy or negotiate agreements based on where we feel costs are set to increase. This can also be used to budget more accurately or with more information.
Project success will be determined based on your ability to identify and provide empirical-based spend/pricing recommendations using models or analysis.
Academic and industry partnerships bring together the best of industry-specific and cutting-edge business knowledge to craft new and innovative ideas.
Mike Pettit
Senior Vice President and Chief Financial Officer, Wabash
