Mutual benefit of cloud manufacturing system and customers through integration of scheduling, order acceptance and fairness under a maintenance strategy

Why smarter shared factories matter

More and more products, from custom phone cases to medical parts, are made in flexible networks of factories instead of a single plant. In cloud manufacturing, companies share machines through an online platform, a bit like renting computing power from the cloud. This raises a big question: how can such a system keep both customers and factories happy at the same time—delivering on time, charging fair prices, and still making a profit—while also dealing with machine breakdowns? This study tackles that challenge head-on.

How shared production in the cloud works

In a cloud manufacturing system, many customers send in jobs to be produced by a set of geographically scattered factories, each with different machines and abilities. The platform must decide which orders to accept, which factory and machine should handle each job, and in what sequence they should be processed. It also has to consider shipping times and penalties for finishing too early or too late compared with the customer’s requested delivery date. Because capacity is limited and missing deadlines is costly, accepting every job is neither realistic nor profitable. The heart of the problem is finding a plan that makes the best possible use of shared resources while treating customers fairly.

Balancing price, fairness, and machine health

The authors build a mathematical model that brings several decisions together instead of handling them separately. First, they set product prices using a “mutual benefit” rule: the price must be higher than the true production and logistics cost for the system, but lower than what the customer is willing to pay. Second, they define customer satisfaction in terms of how much less a customer actually pays than their personal upper limit, and they introduce a fairness goal that reduces the gap between the best-served and worst-served customers. Third, they calculate the system’s own utility from its profit, comparing it to a best-case benchmark so that customer and system utilities can be weighed on similar scales. These three aims—high overall satisfaction, fairness, and good system profit—are optimized together.

Keeping machines reliable and schedules realistic

A key addition in this work is the explicit treatment of machine wear and maintenance. Machines age as they work and may fail unexpectedly. The model allows different levels of preventive maintenance that can “rejuvenate” a machine to varying degrees, as well as minimal repairs after sudden breakdowns that merely restore operation without reducing underlying wear. Maintenance consumes time and money but reduces the risk and cost of future failures. The model simultaneously schedules jobs and maintenance windows on parallel machines across multiple factories, while tracking how these choices affect failure rates, repair needs, and delivery performance. Penalties for finishing too early or too late are also included, encouraging schedules that align closely with customer due dates rather than simply racing to finish as soon as possible.

Testing smarter decision strategies

Because the combined problem is very complex, the authors use an advanced evolutionary algorithm, NSGA-II, to search for a wide set of “Pareto optimal” solutions—plans where you cannot improve one goal (for example, profit) without hurting another (such as fairness). They compare their full integrated model to several simplified versions that omit maintenance, omit preventive maintenance or minimal repair, or ignore earliness and tardiness. Across a range of simulated scenarios, the full model consistently delivers better trade-offs: higher system profit, lower average customer cost, and smaller differences in satisfaction between customers. In some cases, adding the maintenance strategy more than doubles profit while also reducing customer costs and narrowing fairness gaps.

What this means for customers and providers

For a general reader, the main takeaway is that “smart” coordination in cloud manufacturing can genuinely be win–win. By deciding, in a unified way, which orders to accept, how to schedule them, when to maintain machines, and how to set prices, the platform can keep machines healthier, reduce surprise breakdowns, and match deliveries to promised dates. At the same time, it can charge prices that stay within each customer’s comfort zone while avoiding large inequities in how well different customers are treated. The study shows that thoughtful planning and maintenance are not just technical details: they are central levers for building cloud-based production systems that are profitable, reliable, and perceived as fair by their users.

Citation: Salmasnia, A., Abbaszadeh, M. & Kiapasha, Z. Mutual benefit of cloud manufacturing system and customers through integration of scheduling, order acceptance and fairness under a maintenance strategy. Sci Rep 16, 10350 (2026). https://doi.org/10.1038/s41598-026-40759-1

Keywords: cloud manufacturing, production scheduling, preventive maintenance, fair pricing, multiobjective optimization