With the groundwork complete, or at least underway, it’s time to prove the business value of edge computing, AI and advanced analytics. The following steps can guide your initial project to success:

1. Form a small, motivated team. Often, adoption of data-driven solutions needs an evangelist—someone who really believes that AI can transform the business. Be sure the team includes a management-level sponsor, as well as representatives from other aspects of the business such as IT staff, operations staff and customer-facing employees.
2. Put business value first. Identify a real business problem in search of a solution (instead of investing in technology in search of a problem). What customer experience issues can be improved? How can you boost sales? Can supply chain or logistics problems be addressed? The team can brainstorm on several business problems and score them according to importance to the business’ bottom line and customer experience as well as feasibility. Consider both quantifiable business value such as cost savings and profitability as well as qualitative business value such as better customer satisfaction or enhanced premises security.
3. Set a baseline and key performance indicators (KPIs). With a single, feasible, high-value problem identified, determine where you are now. For example, measurements can be by survey (if customer experience is a factor) or analysis of spreadsheet numbers (how was sales over the last three holiday seasons?). Without a baseline, it will be hard to prove the project provided tangible value. KPIs should also be measurable.
4. Establish a strategy and realistic timelines. Considerations include who will be responsible for which aspects of the project, governance of progress and milestones. Regular reporting up to management will keep everyone in the loop and engaged. One framework you may want to adopt is the Cognitive Project Management for AI (CPMAI), which is the industry's best practice for AI, machine learning and data analytics projects.

5. Provide training and communication. Employees may need skilling up, depending on the project. For example, they may need to learn how to interact with edge devices like cameras and sensors. Also, employees may be concerned that their jobs are threatened by AI. Communication about how AI doesn’t replace humans but rather frees humans for higher-value tasks and creativity can help allay those concerns.

6. Be aware of local and state regulations. If you’re deploying cameras and sensors on-premises that gather data about people—customers or employees—and their behavior, you must adhere to relevant regulations pertaining to information security and data privacy. Anonymization of data is critical, as is being able to prove in an audit that you have managed your data appropriately. Unlike the European Union, the U.S. does not have a single, comprehensive federal law that governs privacy. Each state may have its own laws. One way to keep track of this ever-changing area is to use the International Association of Privacy Professionals’ US State Privacy Legislation Tracker.

7. Bring in the experts. Flying solo isn’t usually the key to adoption of emerging technologies. Work with software and hardware suppliers, system integrators, consultants and other edge computing ecosystem players (like Intel) to bolster your own knowledge. Dialoguing about the right combination of technologies can help you make the right choices and help customize solutions.

8. Conduct a pilot project. Keep it small but realistic. For example, you could try out intelligent digital signage for merchandising or smart shelves for inventory management in six stores in a particular region. Or deploy an automated drive-through virtual assistant in a few restaurants. It may be tempting—and less risky—to conduct the pilot project in a lab setting. But an insulated proof of concept that is not exposed to real-world challenges will not truly test the value of the project.