Image by jay huang
Transportation Network Simulation

Business Simulation | Logistics | Optimization

High-fidelity simulations of transportation networks allow for a detailed analysis of freight throughput at cross-docking or warehouse nodes, labor requirements at each facility, probable delays on transportation routes and the downstream impact on customer service. The simulations predict the network behavior under various loads and weather conditions, and provide visibility into the network optimization potential. This knowledge of the future behavior enables management and planners to make decisions about changes to the location of the facilities, hiring of seasonal personnel, changes to delivery schedules, and more general network re-alignment. 

Image by C Dustin
Construction Project Development

Business Simulation | Construction | Estimates

Large construction projects are dynamic systems with multiple external inputs and outputs that are difficult to predict. These complicate the calculation of timelines for the project completion and the actual cost of the project. The simulation samples a large number of scenarios to generate the most likely outcomes for each project. It takes into account the labor behavior, supply chain fluctuations, contractor challenges and other elements that could disrupt the smooth flow of building execution. With the help of the simulation, the company can generate a realistic bid with the timeline and cost that attain the required profit margins without compromising the potential to win the bid.

Image by Guillaume Bolduc
Supply Chain Simulations

Business Simulation |  Manufacturing  |  Risk Management

Supply Chain Simulations model the supplier networks, supplier behavior, the transportation nodes, and transit lanes that deliver the production inputs and distribute the production outputs. Stresses or breakdowns along the chain can generate severe effects on the manufacturing performance and the product availability, especially in a globally distributed manufacturing supply chain. Understanding the range of outcomes helps companies plan optimal inventory levels, order times, and supplier selection both for efficiency and risk mitigation.

Image by Caroline LM
Hospital Services Simulation

Business Simulation | Health Care  | Scheduling

The model captures the provider and services relationships from the EMR software. It builds a relationship graph from these historical data and simulates the patient flow. The daily or weekly simulations generate future data that highlight potential scheduling problems, over utilization of rooms, providers, or other resources.



Image by Dean Ricciardi
Machine Vision for Inspection

Research | Robotics | AI

A tier-one automotive supplier needed to automate the visual inspection of injection molded impellers. The parts had a complex repetitive geometry and qualitative defects that rendered traditional vision systems incapable. The Ordinal team developed AI inspection algorithms that outperformed human inspectors.

Image by Charlein Gracia
Analysis of Childhood Programs

Research | Mathematics | Data Science

A non-profit organization engaged Ordinal Science to analyze a merged multi-agency dataset to explore the impact of childhood programs on delinquency. We developed Markov models to derive a risk score to study the impact of programs on mitigating children's risks of exposure to the juvenile justice system.

Image by Indira Tjokorda
Accurate Real-Time Location Arrivals

Data in Motion | Mathematics | Data Integrity

A transportation company experienced poor accuracy of tracking vehicle arrivals to locations. Spotty GPS data, inaccurate geofences and addresses, poor driver compliance, and incorrect route assignments stymied efforts to collect the key data. Ordinal developed a real-time probabilistic model that merged streams of disparate data achieving 95% accuracy in incomplete data scenarios - a two-fold improvement!

Image by Amanda Easley
Yield Analysis and Combine Calibration

Data in Motion | Mathematics | Data Integrity

A large agricultural concern specializing in precision farming required algorithms to mitigate combine calibration errors. In large farming operations with multiple combines such errors produced erroneous crop yield maps. The Ordinal Science team used kriging techniques in combination with other probabilistic modeling to develop in-line calibration using multi-combine data, significantly improving the accuracy of reported crop yields.

Image by Kai Dahms
IoT Stream Inline Data Correction

Data In Motion | Software

A software company received a stream of sensor and odometer data from telematics devices connected to engines. The odometer was extensively used for generating compliance and business reports, but the values came from too many sources and lacked accuracy. The Ordinal Science team implemented a correction and adjustment model to guarantee the accuracy of the odometers on arrival.

Image by Jessica Lewis
Personnel Scheduling Model

Mathematics | Data Science | Software

A transportation company switched to a new service model that required dynamic scheduling of drivers. A number of complexities with dynamic shipments and shipment volumes caused errors in planning the proper number of drivers at depots throughout the day. The company hired Ordinal Science to develop an optimized model for driver scheduling based on historical data and real-time shipment volume analysis.

Image by Linus Mimietz
Computational Fluid Dynamics

Research  |  Applied Mathematics

An industrial client needed to understand the root cause of sudden breakdowns in their powder production atomization process. Ordinal Science proposed to build a CFD simulation of metal flows over the atypical geometries present in the process in order to explore the conditions that caused the breakdown in controlled atomization. At the end of the project, our team identified the vibration spectrum that caused the breakdown, which could now be mitigated.

Industrial Robotic Arm
Robotic Kinematics Engine

Research  |  Robotics  |  Mathematics

A technology company developing an industrial automation solution utilized existing collaborative robots to manipulate parts in three dimensions. The robot movement needed to be flexible, collision free and smooth in busy 3D space. The existing control algorithms did not perform well in this scenario. The Ordinal team developed kinematics algorithms to satisfy the requirements and built the control library to feed joint positions to the collaborative robot.