Stage 3: run
This module guides testing and deploying automation solutions. We help you meet your organisation’s business needs and realise planned outcomes whilst maintaining legal and ethical compliance.
Outcome-based actions
In Stage 3: run includes testing and building the alpha and beta versions of the solution. Your goal is to implement the outcomes identified in the prioritised opportunities. Implementing change management is critical for automation success, ensuring smooth adoption, transparency, effective stakeholder management, and seamless integration of new processes. Involving users, addressing their concerns, and providing training helps build a positive workplace culture.
During this stage, your automation solutions undergo rigorous testing and validation before being deployed into production. Testing ensures automation solutions meet the organisation’s quality, performance, regulatory and legal administration standards. Your automation team should know the software components including integrated AI, and their value to the organisation. This knowledge enables continuous quality assurance, integration, risk management, scalability, performance monitoring, and optimisation, ensuring the solution’s fairness, reliability, and safety.
In Stage 3: run, your automation team deploys workflow orchestration tools to manage end-to-end process flow after going live and balance the human-bot workforce, developing comprehensive training resources and frequently used automation templates to upskill and educate business users.
Providing business value with an automation software
In simple terms, automation suites are software solutions that use a server-agent model to scale and execute tasks in parallel. A server, or controller, is the core component that manages and orchestrates multiple distributed agents, or bots. Expand below to learn more about some components of an automation software and their business value.
| Component | Business Value | Privacy & Security Value |
|---|---|---|
| Automation engine and robotic process automation (RPA) tools | Enhance operational efficiency by automating routine, rule-based tasks, and processes, reducing manual effort and minimising errors. | Ensure data privacy by automating processes without exposing sensitive information. Security measures include access controls and encryption of data in transit and at rest. Adheres to privacy regulations and industry standards. |
| User interface (UI) and workflow designer | Enable intuitive design and customisation of automated workflows, empowering users to create and modify processes. | Provide role-based access controls (RBAC) to limit access to sensitive workflows. Ensure data privacy through secure handling of user interactions. |
| Integration capabilities | Facilitate seamless integration with existing systems, databases, and third-party applications for end-to-end process automation. | Ensures secure data transfer and integration through encryption and compliance with data protection standards. |
| Artificial intelligence and machine learning integration | Enhance automation capabilities by adding cognitive functions, such as decision-making and natural language processing. | Ensure responsible AI practices, including secure handling of training data, ongoing performance monitoring, and compliance with privacy regulations. |
| Analytics and reporting dashboard | Provide real-time insights into automation performance. It can report on task execution, identify bottlenecks, and optimise processes, allowing data-driven decision-making | Ensure that analytics and reporting tools adhere to privacy regulations and protect sensitive business intelligence. Consider the data collected, stored, and retained to maintain compliance and safeguard privacy. |
| Security and access controls | Mitigate security risks by implementing robust access controls, authentication, and authorisation mechanisms. | Safeguard sensitive data and ensures that access is restricted based on roles and responsibilities. |
| Audit trail and compliance monitoring | Maintain a detailed audit trail of automation activities to ensure transparency and compliance. | Facilitate adherence to data protection regulations by recording and monitoring data access and processing activities. |
| Governance and compliance framework | Establish policies and procedures to ensure that automation aligns with regulatory requirements and organisational goals. | Include measures to enforce privacy regulations, data protection laws, and industry-specific compliance standards. |
| Collaboration tools and notifications | Facilitate communication and collaboration among users involved in automation processes. | Ensure that collaboration tools adhere to privacy standards, and notifications are securely delivered. |
Implementation: human-to-bot balance
Automation suites provide bot development tools that support technical developers and business users. RPA bot development can be script-based (scripted if/then rules) or script-less (Low-code/Drag-n-Drop UI rules). Regardless of how you develop bots, automation teams must cautiously consider the best bot deployment mode for the implemented use case during the build stage. RPA bots run in either attended or unattended mode. Both approaches differ architecturally and operationally. Therefore, technical experts should consider each approach's implication regarding credentials, resources, management, access management, and triggers.
Bot-to-human implementation modes
Description
Triggered and supervised by a human in a process to support triage, approval, or hand-off scenarios.
Useful for
Providing on-the-spot support, enhancing worker’s efficiency and service quality and handle tasks with varying inputs.
Not useful for
Accessing highly sensitive confidential data or a task that has highly variable, dynamic and unpredictable work, and low ROI.
Example
Front-office tasks, customer service, data entry, data migration and clean up, and standard report generation.
Description
Programmed to manage entire process or task independently.
Useful for
Highly repetitive, rule-based, and no human judgment requirement.
Not useful for
Critical and high-stake tasks that requires human empathy, real-time interaction, and judgement.
Example
Back-office tasks, 24/7 operations, non-sensitive community inquiries, case management support, document review and approval.
Unattended bots
It is essential to balance between bots and human service and carefully assess the nature of tasks before deciding on the best-fit bot deployment. Unattended bots should not be deployed to critical high-stake services that mandate human oversight, judgement, and intervention such as emergency response, healthcare, life or death processes. While bots can support professionals with administrative tasks and data analytics, human-in-the loop (HITL) must be maintained in such use cases.
Lack of human judgement
Lack creativity and adaptability
Legal and ethical implications
Lack empathy
Potential for errors
Deployment: scalability, performance and resilience
Automation software servers can be single-node or multi-node. 25,26 Single-node servers are useful for proof of concept (PoC), testing, and demos, while multi-node servers are useful for production with disaster recovery (DR) or high availability (HA) sites. Multi-node servers offer load-balancing, distributing workload across ICT infrastructure. This supports high performance, resilience and service continuity, critical for community-facing services.
The decision-makers and subject matter experts in your organisation should consider the available options for automation deployment and execution. Factors such as security, scalability, compliance, resource availability, budget, and business requirements will impact the environment.
Properly implemented cloud-first deployments provide scalability for business services, especially during peak periods. Many organisations are moving from on-premises toward cloud-based and hybrid deployments to take advantage of cloud scalability, disaster recovery, cost efficiency, and agility while maintaining control over critical data and processes.
In the table below, find deployment and execution options for automation software environments, which can be used individually or in combination.25,26
Description
Installed and managed on the organisation’s servers.
Use-case
- Commonly used for organisations with strict security, regulatory and compliance requirements, e.g., healthcare and finance.
Advantages
- Full control over data and infrastructure
- Regulatory and security compliance
- No external provider services required.
Disadvantages
- Requires significant HW and IT resources
- Limited scalability compared to cloud-based options
- Organisation responsible for maintenance and SW updates.
Description
Deployed on cloud platforms, e.g., AWS, Azure, Google Cloud.
Use-case
- Organisations with variable workloads.
Advantages
- Scalability, flexibility, and remote access
- Cost-effective pay-as-you-go pricing model.
Disadvantages
- Public cloud data security
- Cloud service provider dependent
- Ongoing costs can add up over time.
Description
Combines on-premises and cloud-based elements.
Use-case
- Organisations with hybrid data requirements
- Maintaining data on-premises while leveraging cloud resources.
Advantages
- Data security and control for sensitive information
- Scalability and flexibility of the cloud.
Disadvantages
- Complexity of managing both environments
- Potential integration challenges.
Description
Automation suite components run within virtual machines (VM).
Use-case
- Testing and development environments
- Isolated automation tasks.
Advantages
- Isolation and portability of VMs
- Efficient use of server resources
Disadvantages
- VMs updates and maintenance management
- May not be as scalable as containerisation or cloud-based options.
Description
Automation suite components are containerised.
Use-case
- Microservices-based automation architectures
- Rapid deployment and scaling.
Advantages
- Portability and flexibility of containers
- Ideal for microservices-based automation
- Efficient use of resources.
Disadvantages
- Containerisation technologies skills required
- Additional management overhead for container orchestration.
Description
Automation functions are executed in response to events without managing server infrastructure.
Use-case
- Event-driven automation tasks with variable workloads
- Real-time or low-latency automation requirements.
Advantages
- Eliminates server management overhead
- Cost-effective, as you pay only for actual usage.
Disadvantages
- Limited control over underlying infrastructure
- May not be suitable for all automation scenarios.
Description
- Automation suite is hosted and maintained by a third-party vendor
- Vendors offer management and support.
Use-case
- Organisations seeking ease of management and maintenance
- Reduced IT overhead.
Advantages
- Simplified management and maintenance
- Vendor provides support and updates.
Disadvantages
- Limited customisation and control compared to self-hosted solutions
- Typically, higher costs associated with managed services.
Description
- Automation suite deployed on-premises but accessible remotely
- Offers security and flexibility.
Use-case
- Organisations with sensitive data that need remote access
- Balancing security and flexibility requirements.
Advantages
- Security of on-premises deployment
- Flexibility of remote access.
Disadvantages
- Requires adequate remote access infrastructure and security measures
- Latency issues may surface when accessed remotely.
Description
- Automation suite is deployed on mobile devices
- Suited for mobile automation scenarios.
Use-case
- Field workers and employees requiring mobile automation capabilities.
Advantages
- Direct automation on mobile platforms
- Supports fieldwork and mobile use cases.
Disadvantages
- Limited for automation tasks that require desktop resources or specific integrations
- Mobile device limitations apply, e.g., screen size, processing power.
Description
Automation tasks executed on devices near the data source (edge).
Use-case
- Real-time or low-latency automation tasks
- Situations where processing data locally is critical.
Advantages
- Low latency for critical automation tasks
- Reduced dependence on centralised data centres.
Disadvantages
- Limited processing power and resources compared to data centres deployments
- Requires edge infrastructure and maintenance.
Next lifecycle stage
Stage 4: maintain
Stakeholders are engaged to manage, monitor, and manage automated workflows.