How to Automate Operations Without Breaking the Bank

Another benefit of stand-alone solutions is that many can be adopted incrementally. Companies can start with a simplified solution and later add capacity, or adopt full solutions one at a time and eventually interconnect them. For example, Amco started with a limited set of VNA trucks, but the solution was designed to make it easy to add new sensors and trucks if needed, and the off-the-shelf IoT sensors could share information with other systems if the company eventually needed to create a more interconnected solution. Provided that modularity and compatibility are considered at the outset, a company can adopt a basic version of stand-alone automation and scale it as resources allow and future opportunities arise.

Many AI applications fall into the category of flexible, stand-alone solutions that can be used for operations automation. For example, an SME that provided local logistics services to bottling company Coca-Cola Femsa used an AI chatbot to automate its workforce planning tasks.6 The AI-driven chatbot’s all-in-one functionality allowed it to receive data directly from the user, handle the planning problem by itself, export the results, and even share those results with other company systems through its available APIs, when necessary. As a result, the SME reduced errors in resource allocation and enhanced its efficiency.

Assessing the Cost Implications of Automation Factors

Green circles indicate where low or high levels of each factor lead to the most cost-effective automation solution. For example, a lower reliability factor is sufficient for noncritical activities and is associated with lower deployment costs. Modest reliability and customization requirements are most important when considering a peripheral automation strategy; low integration requirements, high compatibility, and high modularity are most important when considering a scalable stand-alone automation strategy.

A Framework for Cost-Effective Automation

As SMEs set out to pursue the strategies described above, they will need to evaluate specific use cases and will be required to implement them. To help with this, we’ve developed a framework that emphasizes the factors that affect the costs of the proposed strategies. (See “Assessing the Cost Implications of Automation Factors.”) The framework serves as a visual guide that highlights the key cost-related automation factors outlined below.

Reliability. When investing in an automation solution, identify the criticality of the process you intend to automate. Core processes require a high level of automation reliability; hence, the automation solution needs to perform flawlessly for long periods, which can increase costs. Peripheral processes, like monitoring operations, are more fertile ground for low-cost automation because of the lower demands for reliability.

Customization. Tailored solutions may be needed to provide the necessary functionality. Avoid investing in excessive customization unless the automation solution is deployed at a highly critical or responsive operation.

Integration. Prioritize automation solutions that do not require significant integration with IT infrastructure (in other words, stand-alone solutions). This will reduce the implementation complexity, staff time, and effort that needs to be dedicated to the development of an automation solution.

Compatibility. Ensure that your stand-alone solutions can communicate with each other and that their components can be effortlessly combined. Focus on widely accessible technologies with standard interfaces (such as Wi-Fi protocols, APIs, standard data formats, and standard wiring connectors). This decreases the cost of enabling future technology solutions to work together and increases flexibility for adapting the operations.

Modularity. Keep in mind that a rigid automation system (that is, one with all components tightly connected) is difficult to scale and adapt. Build modularity by using technology components that can be replaced or upgraded without severely disrupting the whole process.

For each of these factors, the framework highlights the path to lower automation costs, as represented by the red line in the figure. In this regard, the framework helps identify trade-offs between cost and functionality.

Assessing Automation Costs and Trade-Offs

Let’s look at an example of how to use the framework as a tool to determine whether a solution can be designed or acquired at a low cost. A medium-sized logistics service provider in Boston was interested in implementing a condition-based predictive maintenance solution. The solution was aimed at detecting anomalies in truck component performance, thereby enabling maintenance checks before a breakdown could occur. The company received a proposal from an automation provider for a solution involving sensors and a proprietary algorithm. The logistics provider used the framework to define its requirements and compare them against what the vendor offered, considering each factor in turn.

• Reliability: The company set the level of reliability required at a medium level because some downtime could be tolerated. The vendor solution offered 24/7 monitoring, a higher level of reliability than the company needed.
• Customization: While different vehicles might require different sensors, the company wanted a common algorithm and dashboard across the fleet, so only a medium level of overall solution customization was needed. The provider’s solution, however, varied the feature set by vehicle type, meaning adoption would impose more customization than the company required.
• Integration: The company was seeking a stand-alone solution — one that would, at present, interact only with its vehicles and end users. Consequently, the use case called for a low level of integration, but the vendor offered a solution that required full integration with corporate systems.
• Compatibility: The company considered it likely that, in the future, the solution would need to exchange data with other systems that track additional aspects of vehicle usage, so having compatible components was very important. The solution offered used widely adopted, open IoT standards, aligning with this requirement by easing future data exchange.
• Modularity: Because the fleet was expected to grow, the company wanted a highly modular configuration so it could easily expand or upgrade the solution. But the proprietary solution was seen as having low modularity, with costs mounting as the company added more assets to monitor.

Two important findings emerged from this analysis. First, the required levels for most of the automation design factors corresponded to a low or moderate cost (the green circles in our figure). This suggested that the solution could be designed and implemented internally at a reasonable cost if the company opted for this option. Second, the attributes of the solution offered by the provider either exceeded what was needed or fell short of some requirements the company considered important. That insight enabled the SME to steer its negotiations with the provider, ultimately getting to a lower price by eliminating unnecessary features and selecting those that were truly essential.

We anticipate that the digital divide between SMEs and large companies will continue to grow as the latter heavily invest in automation and AI. It’s important that SMEs understand how they can tap automation technologies affordably in order to maintain competitiveness. If they do so, they will be better positioned to seize new opportunities, improve efficiency, and secure their place in the evolving digital economy.