The future of mining and supply chain automation

Each of the following visions are describing the early majority market segment that represents the entire combined industry supply chain of extracting raw resources, processing and transporting them, and manufacturing them into goods and infrastructure.

20 Years

All operation-level decisions will be automated using advanced AI. Every aspect of the physical process will be monitored using sensors and tracked using an evolving set of metrics and operational objectives. At around this point, both the physical and software aspects of the process will begin to organically evolve and automatically improve itself through experience and adaptation. Humans are only involved in the most niche areas that require specialized knowledge, or high level strategic decisions, and are highly augmented by AI assistance.

10 Years

Many aspects of the physical process are automated but most operational level decisions are still being made by humans. The clear majority of the physical process is monitored by sensors, and advanced machine learning is utilized to optimize processes with the highest return on investment. Many human workers are still utilized in the areas where automation will provide the lowest relative return on investment and where unpredictable market forces prevent investment. All human-driven decisions are made using real-time operational information as well as machine-learned suggestions based on both current and vast amounts of historical data, including data from other operations.

5 Years

Many aspects of the physical process are managed by humans. Tactical decisions are made using relevant, real-time operational information. Some of the most profitable processes are bench-marked against objectives in real-time, and machine learning is utilized to try and advance the efficiency of these processes. Many aspects of the processes are monitored using physical sensors and human input on the front lines. Nearly all repetitive decisions that can be programmed using basic heuristics are automated, have plans to be automated, or are currently being programmed.

Transitioning to Full Automation

Today, humans make most of the decisions in the industrial supply chain, especially in resource extraction. Thirty years from now, almost all the decisions will be made by machines.

For an operational decision to be automated, four things are required:

  1. All information relevant to the decision must be collected in real-time and accessible to the entity making the decision

  2. The machine making the decision must be trained on how to filter and processes the information

  3. There must be some way to communicate the decision to whatever people or machines are required to execute it

  4. The machine making the decision must understand what a good outcome is, so that it can learn from its mistakes and improve over time

Companies interested in being on the leading edge of automation should begin investing in these components today, especially the first. Thankfully, investing in these areas come with both short term and long term returns. Given that the long-term benefit is full automation, and the short-term benefits are more effective decisions made by humans.

The challenge that the industry is currently faced with is access to the relevant information required to make good decisions. Digitization, visualization and situational awareness platforms such as ConnectedWorker™ will be at the center of the next technology trend in heavy industries such as resource extraction and construction.

In general, inefficient decision making costs the industrial supply chain trillions of dollars per year. Investments in these areas will produce exponential returns which will only compound over time as more information is digitized, integrated and made accessible at the times and places decisions are being made.

Mobile and wearable technology represent interfaces that enable better real-time information transfer to and from the front-lines. Although these technologies will be instrumental to any automation transition strategy, their investments will be temporary as front-line human workers are eventually replaced by robotics.

Specifically, the transition is envisioned as follows:

  1. Human workers make decisions based off experience and training and very limited real-time information (Current)

  2. Human workers make better decisions with access to all or most relevant information they need, through digital visualization and intuitive interfaces on desktop, mobile or wearable devices

  3. Human workers make even better decisions with intuitive access to all the information they need on digital interfaces, plus suggestions made by machines that have been learning from past decisions

  4. Human workers almost always follow the suggestions made by machines and generally ignore the source information

  5. Machines make the decisions