Bossa Nova Robotics


May 15, 2019

7 Questions to Ask Your Next Inventory Data Solution Provider

The retail industry is at a critical inflection point. The biggest brick-and-mortar stores are facing significant external pressures, and today’s customers have come to expect a seamless omnichannel shopping experience, every time. With the plethora of new technologies vying for the attention of retailers everywhere, selecting the best solution for your stores can be a daunting task.

We sat down with Sarjoun Skaff, Co-Founder and CTO of of Bossa Nova to discuss why you need to ask these 7 essential questions of your next inventory data solution provider.

Whether you’re compelled to add a mobile robot, drone or fixed camera, these questions will ensure you select an inventory data solution that’s right for the growth of your business - operationally, financially, and technologically.

Sarjoun has over 18 years of experience in long-term robot autonomy, AI, and data analytics. He holds a B.S. in Mechanical Engineering from the American University of Beirut and M.S. and PHD degrees in Robotics from Carnegie Mellon University.

1) How will your data solution address my store's needs, now and in the near future?

Sarjoun: Different data capture techniques, from high resolution cameras to 3D vision, have bearing on the kind of data you receive and the actionable information you can extract from it. It’s critical you understand the current state and future trajectory of your store to select the solution with the technical specs best suited to your space.

Safety is also a factor to consider for technical requirements. Your chosen solution should be able to detect obstacles both low and high and understand how to safely work around them. Ideally, this solution should be able to distinguish between humans, mobile obstacles (e.g. carts), and fixed obstacles (e.g. columns). It should also be strong enough to withstand human meddling, like vandalism or collisions.

2) What data will I receive and when? How do I access this data?

Sarjoun: Out-of-stocks, on-shelf items requiring correction, missed price changes, or missing flag tags are just a few examples of data points to ask for. Establish data delivery timeframe, format, and access permissions at the start to guarantee you get the data you need, when you need it. Also, it is important to confirm that your data has end-to-end encryption. You should be confident that no one else can access your data.

3) Should I use cloud or edge compute to process and analyze my data?

Sarjoun: The cloud offers great economies of scale for data processing, which a serious inventory solution will take full advantage of. Keep in mind that cloud and network outages are a fact of life but should not halt store operations. The ideal solution relies on both cloud and edge processing. The right architecture provides redundancy in the cloud and compute in the store (on the edge) to keep the data flowing in tough conditions.

A word on edge processing - it may be hosted by the service provider, such as in a robot or a docking station, or by the store itself via on-premises compute. Investing in on-premises infrastructure means investing in the future. It can easily be shared by multiple service providers to reduce overall costs, and future technologies that require low latency (such as Augmented Reality) could be more easily realized for shoppers and associates.

4) Can I change the way I operate the store?

Sarjoun: Data has the power to greatly increasing the efficiency of running a store by transforming schedule-based processes into event-driven task allocations. However, such a big change requires an equally big confidence in the data's accuracy and availability. You must:

  • Audit the data over time, in different parts of the store, and across multiple stores to confirm that it is consistently accurate even in varying environments
  • Analyze how long an interruption your operation can tolerate and fix it as an SLA requirement
  • Verify that the right cloud and network architecture is in place for high availability, observability, recoverability, and resilience. For example, if the network goes out, will enough data continue to flow so I can restock my shelves? Is it robust enough to maintain system uptime? What if the data center has a prolonged outage?

Understanding the above will give you the confidence to discontinue old processes, trust the data, and realize its ROI.

5) How long does it take to bring up a new store?

Sarjoun: Installing any new technology can take weeks, if not months. Pick a solution that’s ready to scale, whether you’re planning to add it to 50 or 500 stores. Additionally, after a store bringup, ensure that the data you receive is the data that you want. At Bossa Nova, we are laser focused on getting a store up and running quickly. After an initial training phase performed once for an entire chain, it takes us less than a week to go from zero to daily data delivery for every new store regardless of size and location, and we are constantly reducing that time.

6) Do I need to make any structural changes to my environment?

Sarjoun: New technologies should work around the customer, not the other way around. Consider the implications of having to change flooring, fixtures, lighting etc. An ideal solution should require minimal change to your store environment, if any.

In addition, scanning technologies should not rely on your data to work - such as seeding algorithms with planograms, blueprints or image libraries - as they make compliance circular and are often a source of inaccuracies. Instead, scanning should deliver “ground truth” against which your data can be checked and compared.

7) Are we purchasing the services or the technology itself?

Sarjoun: Traditionally, technology was purchased as capital assets. However, we’ve found that a managed service is a better model for most retailers. At Bossa Nova, you pay for the data and leave the ownership, operation, and maintenance to us. This means you should expect your service provider to build and deploy the new tech; set up new stores; perform remote monitoring, servicing, maintenance and upgrades; and analyze data and deliver actionable information. Determining the long-term service model that is right for your business is just as important as implementing the latest technology.

Whether you are compelled to add mobile robots, drones, or fixed cameras to your stores in 2019, use these questions to evaluate and select a partner that can deliver a viable, scalable, future-proof inventory data service.