Mingzhou Jin

Capacity reservation contracts for high-tech industry

Abstract Capacity reservation provides a risk-sharing mechanism that encourages a manufacturer to expand its capacity more. We propose a deductible reservation (DR) contract where customers reserve future capacity with a fee that is deductible from the purchasing price. The manufacturer’s ex ante announcement of the “excess” capacity that she will have in addition to the reservation amount is a unique feature of the DR contract. An individually rational DR contract that provides channel coordination always exists. Since there is a unique Nash equilibrium for the reservation game among multiple customers, the main results of the one-customer case can be extended to the n -customer case. The DR contract is compared with another capacity reservation contract called take-or-pay. While the manufacturer may gain more profit under a take-or-pay contract, there may not be a channel-coordinated contract that is also individually rational for the customer. Finally, the similarities and differences between the capacity reservation contracts and other well-known supply contracts are discussed.

A column generation approach for the split delivery vehicle routing problem

A column generation approach is presented for the split delivery vehicle routing problem with large demand. Columns include route and delivery amount information. Pricing sub-problems are solved by a limited-search-with-bound algorithm. Feasible solutions are obtained iteratively by fixing one route once. Numerical experiments show better solutions than in the literature.

A case study of exploiting enterprise resource planning requirements

The requirements engineering (RE) processes have become a key to conceptualising corporate-wide integrated solutions based on packaged enterprise resource planning (ERP) software. The RE literature has mainly focused on procuring the most suitable ERP package. Little is known about how an organisation exploits the chosen ERP RE model to frame the business application development. This article reports an exploratory case study of a key tenet of ERP RE adoption, namely that aligning business applications to the packaged RE model leads to integral practices and economic development. The case study analysed a series interrelated pilot projects developed for a business division of a large IT manufacturing and service company, using Oracles appl1ication implementation method (AIM). The study indicated that AIM RE improved team collaboration and project management experience, but needed to make hidden assumptions explicit to support data visibility and integrity. Our study can direct researchers towards rigorous empirical evaluations of ERP RE adoption, collect experiences and lessons learned for practitioners, and help generate more effective and mature processes when exploiting ERP RE methods.

Cross-Facility production and transportation planning problem with perishable inventory

This study addresses a production and distribution planning problem in a dynamic, two-stage supply chain. This supply chain consists of a number of facilities and retailers. The model considers that the final product is perishable and therefore has a limited shelf life. We formulate this problem as a network flow problem with a fixed charge cost function which is NP-hard. A primal-dual heuristic is developed that provides lower and upper bounds. The models proposed can be used for operational decisions.

A Systems Approach to Product Line Requirements Reuse

Product line engineering has become the main method for achieving systematic software reuse. Embracing requirements in a product lines asset base enhances the effectiveness of reuse as engineers can work on the abstractions closer to the domains initial concepts. Conventional proactive approaches to product line engineering cause excessive overhead when codifying the assets. In this paper, we propose a systems-oriented approach to extracting functional requirements profiles. The validated extraction constructs are amenable to semantic case analysis and orthogonal variability modeling, so as to uncover the variation structure and constraints. To evaluate our approach, we present an experiment to quantify the extraction overhead and effectiveness and a case study to assess our approachs usefulness. The results show that our automatic support offers an order-of-magnitude saving over the manual extraction effort without significantly compromising quality and that our approach receives a positive adoption rate by systems engineers.

Multiple unit auctions with economies and diseconomies of scale

Abstract This paper discusses multiple unit auctions for industrial procurement where the cost structures of suppliers capture economies and diseconomies of scale caused by the nature of the production cost and the opportunity value of suppliers’ capacities. The problem of winner determination and demand allocation is proven to be NP-complete. We propose a binary tree algorithm with bounds (BTB) which efficiently exploits the model’s optimality properties. BTB outperforms general integer optimization software in computational time, especially with existence of substantial economies and diseconomies of scale. The algorithm complexity is linear in demand volume. This property allows for efficient handling of high volume auctions and thus leads to increased benefit for the overall system. Under the assumption of the myopic best response strategies, we investigate the behavior of suppliers and price dynamics for iterative (multiple round) bidding with appropriate allocation and stopping rules. The allocation rules, featured by several tie breakers for multiple optimal solutions to the allocation model in each round, are proposed to induce suppliers’ dominant strategies and to improve the system’s performance.

Optimal routing of vehicles with communication capabilities in disasters

Major emergencies and disasters such as acts of terrorism, acts of nature, or human-caused accidents may lead to disruptions in traffic flow. Minimizing the negative effects of such disruptions is critical for a nation’s economy and security. A decision support system that is capable of gathering (real-time) information about the traffic conditions following a disaster and utilizing this information to generate alternative routes for vehicles would benefit the government, industry, and the public. For this purpose, we develop a mathematical programming model to minimize the delay for vehicles with communication capabilities following a disaster. Most commercial trucks and public buses utilize QUALCOMM as a communication tool. We also develop a prediction model for vehicles that do not have any communication capabilities. Although the problem is inherently integer we developed a linear program to reduce the computational burden caused by the large size of the problem. An algorithm is proposed to update the parameters of the linear program based on a duality analysis in order to obtain better results. A monotonic speed–density relationship is embedded in the model to capture high traffic congestion that occurs after a disaster. The model and the algorithm are tested using a simulated disaster scenario. The results indicate that the proposed model improves system performance measures such as mobility and average speed.

EZ M-Parking System

M-commerce has been applied in parking payment systems since 2001. However, its complicated user interface increases time for motorists to complete transactions and, therefore, hurts the popularity of mobile phone-based parking systems. A new m-commerce parking system called the EZ M-parking system is proposed for both on-street and off-street parking. A unique number including a check digit is assigned to and painted on each parking space. Motorists need dial only a single number to start or end their parking service. Short message service (SMS) messages are used to send confirmations or reminders to motorists. Parking enforcement officers can check the occupancy of a parking space online or based on the SMS messages received from the management center. Officers can also send SMS messages to motorists if necessary. This system allows flexible pricing and accurate charging and includes several methods to reduce operation errors. The real-time parking occupancy data can help develop parking guidance system...

Dynamic railcar connection planning in classification yards

This paper investigates dynamic railcar planning in railroad classification yards. The plan decides the assignment of railcars from inbound trains to outbound trains under various size limitations of outbound trains and allows dynamic sequencing of inbound train classification and outbound train assembly. A mixed-integer program is presented for the problem along with a heuristic algorithm based on the harmony search strategy. The numerical results based on both a real-world case and randomly created instances show that dynamic sequencing can reduce total dwell time compared to static sequencing, especially when more trains are classified at a yard. However, the savings on average dwell time for each railcar become smaller when the traffic volume goes up. Once the volume through a yard is close to its capacity, the dwell time per rail car goes up very quickly and the benefit of dynamic sequencing diminishes. It implies that the benefit of dynamic sequencing is more significant for yards with lower traffic volume.

Sequencing and Scheduling in Railway Classification Yards

Three sequencing decisions are involved in railway classification yards: inspection, hump, and assembly. Those sequencing decisions are important for smooth and efficient railyard operations and are interrelated. This research concentrates on the optimization of those sequences in an integrated way. A sequencing and scheduling model is proposed by considering multiple engines and inspection groups. This model can be solved by existing commercial optimization solvers for one typical planning horizon, such as 24 h. Numerical experiments and a case study based on historical data from a U.S. Class I railroad demonstrate that the proposed solution method yields better sequences and schedules, as measured by the total dwell time, compared with the practice of static sequencing. Furthermore, the results indicate that the handling capacity should be balanced among different classification steps to maximize the overall yard capacity.