Mining and energy companies face some of the most significant challenges in driving internal collaboration - particularly when it comes to realising data-driven projects.

Some staff are on-site, and even underground, some are in the middle of the ocean, while others are distributed between home and office towers. Cross-department team members are scaling projects of such size and complexity that garnering consistent and evolving buy-in from the wider business can quickly become arduous (read: expensive). And let’s not forget, the distances between team members can be hundreds, if not thousands of kilometres - and that’s across the ground, it doesn’t even scrape the distances hidden beneath it.

A breakaway from the silo mindset is one of the most significant challenges that innovative teams need to conquer to find organisation-wide success, especially when it comes to machine learning solutions. Here, I explore how big business is tackling segregation between departments, and how that’s ultimately going to enhance the application of AI and machine learning products in future.

Silos leave innovation in the lurch. Why?

A silo mentality within an organisation can emerge for several reasons; perhaps senior management isn’t aligned on the business’ core values and strategic objectives, are seeking to undermine their counterparts, or, maybe the mere fact that multinationals are spread over such distances, and across such complex knowledge bases, it just doesn’t seem convenient or practical to keep information free-flowing.

There are incredible opportunities out there for resource companies to undergo a digital transformation: saving lives, avoiding injuries, and reducing CO2 emissions are just some of the key benefits of machine learning’s impact on personnel and the environment. Organisations are now looking towards AI to recover mineral deposits that would otherwise remain concealed, or creating greater efficiencies in their machinery - these are just a couple of scratch-the-surface examples.

The World Economic Forum has projected a $425 billion value to society in the application of machine learning in resources, and $320 billion of value to the industry, in just the next ten years. So why is machine learning such a challenge to get off the ground, internally?

Contrary to sweeping opinion, it can be difficult to identify opportunities for appropriate use cases and difficult again for organisations to recognise the need for a machine learning solution, at all. And, if a successful use case scenario is identified, communicating that identified need becomes an uphill battle over different knowledge bases and through corporate politics - which means the biggest challenge for AI and machine learning in mining and energy is, in fact, human - partly because people don’t trust predictive algorithms and partly because an automated process requires an adaptation to human behaviour.

One approach that is popular for the application of a machine learning solution is to stagger the rollout. For example, here are four stages to machine learning replacing a human decision-making process:

1. Machine learning is applied to problems to understand the causes of past events - for example, a machine failure.
2. Machine learning is applied to predict future failures. Continuing with our machine failure example, this would allow a team to act preventatively.
3. An application of machine learning that aggregates ongoing failure data and continues to predict future instances - our team can now supervise this process both ways and react accordingly to streamline machine maintenance.
4. Machine maintenance becomes fully automated, and there is now no human input on this issue - our example machine management team can direct their focus elsewhere.

The final goals achieved by this sampled automation rollout include cost efficiencies and improved sustainability outcomes; needless and expensive overservicing of machines is no longer a pain point, and the business can redirect those resources elsewhere.

Alas, enter the silo mentality. People become the spanner in the works when they suddenly need to shift from their established behaviours to accommodate new processes and technologies. To retrofit a business with a new machine learning solution can result in some cross-department adjustments and, where communication starts to break down, cost duplications, trust issues, team morale and unity can closely follow.

Let’s set the scene:

You, an Exploration Geologist, work in a large mining company and are facing an optimisation challenge with, say, finding new economic mineral deposits. As the subject matter expert in exploration, you understand full well that there’s an opportunity to try a different approach to solve this problem, however, you’re not quite sure how a data-driven machine learning product - purchased or otherwise - is supposed to optimise your current output.

Your senior management team brings another piece to this puzzle: budget approval. This team is exerting careful governance over budget, and though they desire to create cost efficiencies, they don’t have experience with machine learning products and can’t assess this as an evidence-based solution with their existing experience. Sure, senior management thinks that optimisation to the current process would be excellent but the science and workload behind a prospective customised solution seems overwhelming, even when they’re interfacing with experts. They’re not sure how to attribute value to a solution that hasn’t been born yet. 

Meanwhile, your distant data scientist colleague has the final piece of the puzzle: the digital skills to apply a new approach to the problem. Sadly, having suggested there might be a data-driven solution in the first place, your colleague doesn’t have a.) the authority and b.) the capacity to interface with senior management on your team’s behalf. Assuming our data scientist had been advocating for a new solution for some time, there’s a good chance they have struggled to get the buy-in required - this could be attributed to ‘knowledge silos’ and effective communication strategies between departments or over long distances.

It is important to note that each of our three stakeholders brings a unique skillset and contribution to the overall picture - as separate entities, they have equal power to bring a project to fruition or squander the opportunity. The communication walls are high, there is a knowledge barrier between each participant and, ultimately, you either keep up with a current methodology for the indefinite future or embark on implementing a solution under less-than-ideal collaborative circumstances.

Identifying the symptoms of a silo work culture

The ramifications of a silo work mentality are wide-reaching and potentially devastating, not only for employee welfare but also for a company’s potential - that’s potential revenue, innovations, employee satisfaction, global reputation, and more. Do any of the following apply to your office?

• Is it challenging to meet with, borrow equipment from, or share with/extract information from another department?
• Do employees tend to communicate requirements up the chain of command, instead of requesting assistance directly?
• Are different areas of the workforce treated differently in regards to their uniform expectations, workplace guidelines, and standards of excellence?
• Are decision-makers conservative or non-inclusive when it comes to accepting new ideas from the wider group?
• Does the workforce utilise different operating systems and software applications per department, with no central port to connect?

If you’ve answered ‘yes’ to any of the above, it may be time to consider taking steps towards blasting through some silo structures.

Overcoming silos: How mining and energy staff can collaboratively establish a return on investment

So, when locations segregate team members, departments, projects, skillsets and cliches, what brings them back together? What aligns an organisation towards a common goal?

A shared vision of the future

Giving careful consideration as to how silos evolved in the first place, the initial step towards rekindling flames between departments requires a consolidated vision for the company, moving forward. The classic ‘where do you see yourself in five years’ line of questioning doesn’t just apply to a potential new employee, it also applies to the conception of united company culture. This ‘vision’ needs to identify not only the company’s areas of strength but the areas fostering new growth and overarchingly, the ‘why’ behind the company’s mission statement. Where is the team going, and why?

Uniting a large workforce is no simple task, nor is there a perfect approach to achieving this. Still, I’d say the building blocks for success starts with an identified vision, a mission statement, core values, and openly communicated top-tier goals.

A platform to be heard

Project managers at every level from around the world are embracing software solutions to help teams collaborate. And, with enormous offerings out there, businesses are spoilt for choice when it comes to finding the right tool to communicate, schedule, update, visualise, store, and raise support tickets for projects and teams of all sizes. Implemented correctly, large organisations will thrive - to quote Salesforce:

“A unified vision of company goals will turn damaging silo mentality into a culture where there is a division of labour but not a division of information.”

Collaborative tools for enterprise have transformed the way companies live and breathe all over the world. Though my experience within the mining and energy sector has shown that despite being at the forefront of automating mobile equipment, a traditional working mindset is often prevalent - particularly when it comes to organisational structure and communication pathways from top to bottom, and bottom to top. And it’s not as though big business isn’t aware of these hierarchical issues - most are running regular surveys on their workforce to measure and improve inclusive teams. From my perspective, the right collaborative tool gives both the business and each employee a platform to broadcast, to be heard, to collaborate, and to access the key insights and materials that will drive the business forward.

A reward structure aligned with company objectives

Even a ‘motivated team’ may experience fluctuating morale. In business, a rewards structure can help significantly with getting employee buy-in for organisational change. Imagine for a moment that an employee’s bonus wasn’t tied to his line manager’s KPIs, but instead, rewards were structured around the organisation’s core values and long-term mission milestones. Just think, instead of motivating employees for the short term, the ‘hive mind’ could be collaborating towards goals for the next 5, 10 or even 20 years. By pushing staff outside of their comfort zones, exceptional outcomes can be met.

Read also: Machine Learning & Exploration Data: How to overcome common issues during data preparation

Justin Strharsky is Founding Director of Unearthed solutions, the largest community of startups, developers and data scientists making the energy and resources industry more efficient and sustainable. Justin has previously worked in Silicon Valley, for a large network computing company as well as one of the first companies that commercialised artificial intelligence.