XeroData

XeroData is a software and data consulting company that develops agent-based models of supply chains. Typical supply chain management software uses complex equations based on network flow and dynamic optimization—making it difficult to integrate realistic details of how companies interact with one another. Conversely, multiagent models integrate many kinds of agents (e.g. suppliers, purchasers, wholesalers), and make it far easier to customize the model, understand it intuitively, and calibrate it to real-world data. Especially in industries like oil and gas, the majority of costs come from logistics. Operations managers need to rapidly handle shocks such as equipment failures or world events whose effects propagate throughout the supply network. They also need to establish contracts with transporters, which they still need to pay for even in the event of a disruption. Much data is available, but companies in a network must share it on a centralized platform so that all can benefit. Other issues, such as legal and environmental factors, are well-known within the industry but difficult to account for in traditional mathematical models. Our solution is computer simulations, which are well-known for their effectiveness in handling intractable problems. The traveling-salesman problem of finding the quickest route among cities is the fundamental problem of logistics, but famously has no best algorithm. Yet, simulations perform very well on such problems—from simulated annealing inspired by metallurgy, to swarm intelligence inspired by ant colonies.