Industrial optimisation at the speed of operations

Industry focus

• Mining: mine planning, fleet dispatch, blending, maintenance, and constrained production optimisation.
• Energy: grid-aware scheduling, asset coordination, storage dispatch, and renewable integration.
• Logistics: routing, network planning, yard operations, fleet allocation, and disruption recovery.
• Manufacturing: manufacturing process optimization, production scheduling, setup sequencing, and bottleneck removal.

Quantum-ready computing for industrial optimisation

Hyperbridge maps combinatorial industrial problems into formulations suited to emulated quantum algorithms, quantum annealing, and gate-based approaches while remaining deployable on existing GPUs and FPGAs. This helps teams solve larger optimisation models with more constraints and faster re-planning cycles.

Key pages: About Hyperbridge, How it works, Data and IP, AI summary.

Contact

Customers and partners can book a meeting at cal.com/team/hyperbridge/hyperbridge-dr.

Mining — quantum-ready optimisation for mining operations

The Hyperbridge mining section is organised as a hub with two horizon-specific silo pages. Everything runs alongside the existing planning, scheduling, and dispatch stack — nothing is replaced. Benchmarks below were produced on a single consumer-grade Nvidia RTX 4050 laptop GPU.

• Mining hub: value promise, what we hear from planners, and links to the two silo pages.
• Strategic mine planning: pit shells, pushbacks, schedule, and fleet optimised together in one pass.
• Continuous shift re-planning: re-solve the full shift fast enough to keep the plan and the operation together.

What planners tell us. The plan and the operation drift apart: actuals move away from the plan through the shift as faces, trucks and priorities change, and the re-plan cadence cannot catch up before the gap compounds. Model detail is rationed by solve time. Deep geological and operational expertise often cannot reach a viable recommendation while the decision still matters. Where the bottleneck is compute, Hyperbridge removes it.

Strategic mine planning benchmark. On a representative synthetic 10-year plan optimising pit shells, pushbacks, schedule and fleet together, Hyperbridge reaches an optimal plan in about 10 minutes. An OR-Tools baseline did not converge to optimal within 12 hours. Against that baseline: NPV about 6% higher on the same orebody, peak fleet reduced from 5 diggers / 41 trucks to 4 diggers / 27 trucks (around 30% leaner). Signed proof-of-value engagement underway with an independent mine planning consultancy. See /mining/strategic-planning.

Continuous shift re-planning benchmark. A production-realistic synthetic scenario with 2 pits, 5 dig faces per pit, 40 trucks, 3 crushers and roughly 214,000 variables re-plans the full shift in around 11 minutes. Open-source baselines on the same problem: CBC still 20% off after 24 hours, OR-Tools still 1% off after 12 hours. Sites typically carry around 10% more trucks and diggers than a matched plan needs; re-solving the whole shift in minutes removes the stale-plan share of that buffer. Validated by a serving mine planning architect at a Tier-1 iron ore operation. See /mining/shift-replanning.

Energy — optimising the system as it gets harder to schedule

Every power system solves the same problem repeatedly: which generators and storage to run, and at what level, to meet demand at lowest cost within a web of physical and network constraints. It is NP-hard, and as renewables and storage grow the problem gets bigger, conditions shift faster, and operators must solve it more often and further ahead. The tools that schedule the grid are becoming the limit on how well it can run.

Where we think we can add value. Two markets show the shape of it. In the United States, grid-operator day-ahead scheduling runs to hundreds of thousands of variables against a roughly twenty-minute deadline; operators stop at an optimality gap because proving the best answer in time is not feasible. Closing that gap and extending the look-ahead is where we believe we can help. In Australia, the National Electricity Market dispatches every five minutes, with prices swinging from minus $1,000 to $20,300 per megawatt hour; Western Australia moved to look-ahead, security-constrained co-optimisation in October 2023 with renewables now around a third of supply. Solving that problem quickly and well as it grows is where we believe we can help.

The performance we describe for energy is a hypothesis we want to test, not a measured result. Our benchmarked results to date are in mining. Learn more at /energy.

Logistics — continuous routing when the plan meets reality

Moving goods through a city is the vehicle routing problem — which vehicle serves which stop, in what order, within which time window, under which capacity and traffic constraint. It is NP-hard. Mature heuristics handle a single morning plan well, but the city does not hold still: a driver calls in sick, a road closure removes routes at noon, same-day orders arrive after cut-off, an afternoon spike pushes past planned capacity. Each event invalidates the morning plan.

The window to respond is set by the work itself. A driver is committed to the stop in front of them; the decision that matters is where they go next, and it has to be ready before they complete that drop — a few minutes in dense urban delivery. The real task is a high-quality re-plan of the remaining work delivered inside that gap, again and again across the fleet through the day.

Where we think we can add value. An operator running thousands of vehicles and tens of thousands of stops across a metropolitan area sits exactly where conventional re-optimisation breaks: it returns a worse plan, or a good plan too late. This is the same class of problem as our mining work — a large, constraint-heavy schedule rebuilt inside a tight decision window when something breaks. The performance we describe for logistics is a hypothesis we want to test with a partner, not a measured result. Our benchmarks to date are in mining. Learn more at /logistics.

Manufacturing — optimising the factory schedule as production gets harder to plan

Every factory solves the same problem shift after shift: which job runs on which machine, and in what order, to meet due dates and protect throughput within routing, setup, batching and capacity constraints. In its flexible form the job-shop scheduling problem is strongly NP-hard. As product mixes widen, lines share more equipment, and orders change inside the day, the schedule a planner fixed in the morning is stale by mid-shift.

Where we think we can add value. Three environments show the shape of it. Semiconductor fabs run several hundred operations across roughly a hundred tool groups with re-entrant flow; because re-solving in time is not feasible they fall back on dispatching rules and accept idle tools. Pharmaceutical and fine-chemical plants campaign many low-volume products on shared multi-purpose equipment with cleaning and validation windows. Primary metals couple casting, hot and cold rolling and coating with sequencing rules on width, gauge, grade and temperature. Solving these schedules quickly, and re-solving as conditions change, is where we believe we can help.

The performance we describe for manufacturing is a hypothesis, not a delivered result. Our proven performance to date is in mine planning; the manufacturing problems above sit in the same class of constraint-heavy combinatorial scheduling, which is why we believe the approach transfers. Learn more at /manufacturing.

About Hyperbridge

Hyperbridge is a compute and optimisation company headquartered in Brisbane, Australia. We build an engine that solves large industrial scheduling and planning problems on the hardware operators already own, with a clean path to quantum hardware as it matures.

Founders. Ray Tseng, Chief Commercial Officer — 30 years building and scaling enterprise technology across mining, ports, telecommunications and manufacturing; ex-Microsoft, Nokia, Metaswitch; Carnegie Mellon (ECE) and HKU (MBA). James Donnithorne-Tait, Chief Technology Officer — leads software development, deployment and Hyperbridge's customer proofs of value; ex-Microsoft, Metaswitch; Cambridge Natural Sciences (computational psychology); based in Brisbane. Learn more at /about.

How Hyperbridge works — Encode, Solve, Deploy

Encode. Your operational problem — which trucks, which faces, which order, under which constraints — is translated into a form the engine can solve. The modelling layer carries the real constraints of your operation: equipment behaviour, business rules, and failure modes, so the answer the engine returns is one you can actually run.

Solve. The engine searches the solution space in parallel, the way a physical system settles into a stable state. That is what makes it fast on hard combinatorial problems where conventional solvers slow down as the problem grows.

Deploy. It runs as a software block (C++ / Python) alongside your existing planning and dispatch tools. No redesign of your processes is required. Learn more at /how-it-works.

Why now. GPUs sat in machines for years before a software layer unlocked them for general compute and made the modern AI market possible. High-performance optimisation and control have stayed largely on CPUs. Hyperbridge brings that same class of acceleration to industrial optimisation.

What you need. Runs on commodity GPUs today (our mining benchmarks ran on a single consumer-grade Nvidia RTX 4050 laptop GPU); scales to accelerator hardware (GPU, NPU, FPGA); the same encoding runs unchanged on quantum hardware as it matures.

Data and IP — your operating knowledge stays yours

What stays yours. Your raw operational data, your site-specific constraints, and your results remain your property and are not shared with any other customer.

How we deploy. Hyperbridge runs alongside your existing stack on-premises, in your cloud, or ours. We work with you to choose the right implementation for your business.

What we retain, and what we do not. Hyperbridge may retain the configuration, workflows, and system-level learnings required to operate, support, and improve the platform for your deployment. We do not claim ownership over your raw data, operational records, site-specific constraints, business rules, production outputs, or proprietary process knowledge. We do not use one customer's data, results, or operational details to train, configure, or optimise another customer's deployment without explicit permission.

Security and confidentiality. Customer data and operational knowledge are treated as confidential business information. Access is limited to authorised personnel and systems required to deploy, maintain, secure, and support your Hyperbridge environment. For customers with specific compliance, data residency, audit, or confidentiality requirements, Hyperbridge can work through those requirements during technical and commercial review. Learn more at /data-and-ip.