Data-driven throughput bottleneck analysis in production systems
Doktorsavhandling, 2021

Production systems and production management are getting smart. Manufacturing companies are increasingly adopting digital solutions to monitor and manage production systems. By adopting digital solutions, it has become possible for manufacturing companies to collect huge volumes of widely varying production system data. Alongside this, significant advances in recent years in machine learning and artificial intelligence fields have opened up opportunities to develop data-driven approaches for analyzing the huge emerging volumes of data, deriving insights, and using those insights to improve shop floor productivity.

One way to increase shop floor productivity is to strive to achieve an even flow within production systems. However, even system flows may be disturbed by various factors such as random breakdowns, minor stops, setups, variations in cycle time, waiting for an operator, and so on. Such disturbances may constrain production system throughput. However, previous research has shown that not all disturbances in different machines in the production system
constrain the production system throughput. The disturbances are significant in a set of machines in the production system that constrains the system throughput. This set of machines is called throughput bottlenecks. Quick and correct identification of throughput bottlenecks will help practitioners plan appropriate eliminating actions.

Existing academic research efforts to investigate throughput bottlenecks have largely adopted analytical approaches or discrete-event simulation-model-based ones. However, now that companies are collecting huge volumes of digital data, this data can be analyzed directly by developing data-driven approaches. This allows the derivation of insights into throughput bottlenecks in production systems.

This doctoral thesis constructs a series of data-driven approaches to analyze throughput bottlenecks. Firstly, a data-driven approach is proposed to identify historical throughput bottlenecks in a production system. But merely identifying bottlenecks is not enough if informed actions are to be taken. Secondly, a data-driven approach is proposed for diagnosing historical throughput bottlenecks. Specifically, a diagnosis is made based on a maintenance perspective. Combined with identification and diagnosis, practitioners may then plan and execute different corrective actions. However, when such corrective actions are applied, the dynamics of the production system change. That means that bottlenecks will not act as bottlenecks in the future. Thirdly, to thus predict how the system dynamics will change (and thereby to predict future throughput bottlenecks), a data-driven approach is proposed to predict future throughput bottlenecks in a production system. Fourthly, to help practitioners plan for proactive actions on the predicted throughput bottlenecks, a further data-driven approach is proposed; one which prescribes actions to combat the bottlenecks.

The different data-driven approaches proposed in this thesis have been tested using production system data sets extracted from different real-world production systems. The insights obtained from applying these approaches may help practitioners to better understand the dynamics of throughput bottlenecks and plan for specific actions to eliminate them. Such elimination helps achieve an even flow in production systems, thus increasing productivity.


machine learning



manufacturing system

production system

artificial intelligence

Throughput bottlenecks

VDL and Zoom (Password: 534528)
Opponent: Amos Ng, University of Skövde, Sweden


Mukund Subramaniyan

Chalmers, Industri- och materialvetenskap, Produktionssystem

A data-driven approach to diagnosing throughput bottlenecks from a maintenance perspective

Computers and Industrial Engineering,; Vol. 150(2020)

Artikel i vetenskaplig tidskrift

A generic hierarchical clustering approach for detecting bottlenecks in manufacturing

Journal of Manufacturing Systems,; Vol. 55(2020)p. 143-158

Artikel i vetenskaplig tidskrift

A prognostic algorithm to prescribe improvement measures on throughput bottlenecks

Journal of Manufacturing Systems,; Vol. 53(2019)p. 271-281

Artikel i vetenskaplig tidskrift

A data-driven algorithm to predict throughput bottlenecks in a production system based on active periods of the machines

Computers and Industrial Engineering,; Vol. 125(2018)p. 533-544

Artikel i vetenskaplig tidskrift

Data-driven algorithm for throughput bottleneck analysis of production systems

Production and Manufacturing Research,; Vol. 6(2018)p. 225-246

Artikel i vetenskaplig tidskrift

An algorithm for data-driven shifting bottleneck detection

Cogent Engineering,; Vol. 3(2016)p. 1-19

Artikel i vetenskaplig tidskrift

This thesis offers a series of data-driven solutions that can provide a multifold increase to a company's bottom line by eliminating throughput bottlenecks in the factory. Today manufacturing companies' factory floor productivity is alarmingly low at 50%. Practitioners are exploring new ways to increase factory floor productivity. One way to increase productivity is to get higher factory throughput. Some machines constraints the throughput on the factory floor. These machines are called throughput bottlenecks. When practitioners eliminate throughput bottlenecks, they can get higher throughput. But how can practitioners find, analyze, and eliminate throughput bottlenecks? Currently, practitioners spend a lot of time (sometimes hours or days) on the shop floor to search for bottlenecks and make ambiguous experience-based decisions. But this can be changed using digital solutions. How? My research answers this question. In this thesis, I build data-driven approaches to analyze throughput bottlenecks in less than seconds. The input to a data-driven approach is digital machine data. Then, the data-driven approach quickly analyses the digital data using artificial intelligence techniques. The outputs of a data-driven approach are the insights on throughput bottlenecks.

Within this thesis, I propose four data-driven approaches for different types of throughput bottleneck analysis. First, I present different data-driven approaches to identify historical throughput bottlenecks. With these, practitioners can quickly identify the bottleneck location in a production system. Second, I propose a data-driven approach to diagnosing historical throughput bottlenecks. It will help to understand the possible root-causes of the throughput bottlenecks. Third, I offer a data-driven approach to predict throughput bottlenecks for the next production day. It will help to take proactive actions on throughput bottlenecks. Fourth, I propose a data-driven approach to prescribe actions on predicted throughput bottlenecks. It will give information on specific measures one can proactively perform on predicted throughput bottlenecks. In sum, these data-driven approaches will help practitioners to make faster, confident, and informed decisions on throughput bottlenecks, which will help to maximize the throughput from production systems.

Overall, data-driven approaches are similar to GPS. People use GPS to find the best way. The GPS eliminates blind alleys. Similarly, practitioners can use the data-driven approaches to eliminate throughput bottlenecks and create a more predictable, safe, and better factory environment without surprises.

DAIMP - Dataanalys inom underhållsplanering

VINNOVA, 2016-03-01 -- 2019-02-28.


Produktionsteknik, arbetsvetenskap och ergonomi





Doktorsavhandlingar vid Chalmers tekniska högskola. Ny serie: 4887


Chalmers tekniska högskola

VDL and Zoom (Password: 534528)


Opponent: Amos Ng, University of Skövde, Sweden

Mer information

Senast uppdaterat