- The ambient IoT technology company built an AI-based system for its own purposes to interact with the data being captured about moving goods.
- Companies that ship and sell fresh products such as seafood are now leveraging the system to “talk” to their products and learn details such as freshness.
Imagine this: a talking crate of seafood comes into a store warehouse. The facility manager has questions for it: “Hey, how were you treated along the way to our facility?” or “How fast do I need to have you on the store shelf?” or even “How fresh are you in comparison to the other pallets in the shipment?”
IoT technology already collects data that could be used to answer those questions, from location of tagged goods to sensor-based details; but with AI, that data can generate logical information provided directly to the user, in conversation.
With that in mind, ambient IoT technology company Wiliot has built an AI-based, chat-bot solution known as WiliBot.
The company already offers an IoT solution that includes wireless tracking of goods and related sensor information. Its IoT Pixels tags transmit Bluetooth Low Energy (BLE) signals to area gateways, and Wiliot’s software manages that data to understand if a product is in transmit, where, and in what condition.
Tsunami of Data
For those collecting the data, it can feel like they are on the receiving end of a tsunami of data, said Steve Statler, Wiliot’s chief marketing officer. Traditional passive RFID systems provide snapshots of what happens in a supply chain: if a product or pallet is tagged, that tag could be read at a fixed reader portal in a dock door, for instance, or with a handheld reader. The Wiliot system collects much more data.
So for companies using such technology (and viewing the resulting data), one challenge can be, “how do I make sense of this and use it for useful purposes that will help reduce food waste, or have better healthcare outcomes and so forth,” said Statler.
Recognizing the same challenge in-house, Wiliot developed its AI tool in May “to help us deal with the problems and opportunities that happen when you get an omniscient view of everything, everywhere all at once,” he stated.
The next step was to make the same tool available to customers.
“We made the decision to move WiliBot from an internal tool to something that our customers and partners can use,” said Statler.
Gaining Clarity about Supply Chain
One typical use case would be the movement of seafood from a dock, through a temperature-controlled warehouse, to a store’s own refrigerated displays. Eric Casavant, Wiliot’s director of technical marketing, pointed out that many problems can occur in the supply chain that, until recently, the constituents were unaware of— goods sitting too long outside of refrigeration, being delayed in transit, or being put out for consumers in the wrong order (with goods that will expire soonest, bypassed by other goods that have a longer shelf life).
One customer has been using WiliBot to better understand their own fish supply chain, through the conversational interface. Other pilots are launching as well, using the AI feature.
Users can leverage the system as an app and start a conversation based on real-time and historical data, Casavant said.
For instance, if Wilibot were asked if a specific product was handled properly, it might respond that there was a four-hour period when it was out of temperature compliance.
The follow up question of “where did that happen?” could lead to a description of the site and time, what other products were affected, how often does this happen at this specific site and on what days and times do these kinds of lapses occur.
Enhancing Expiration Details
The system can be used by operators as well as managers, Casavant added. If a stock person is placing fresh shrimp on store shelves, they could ask WiliBot which product should be placed there first—based on not only expiration dates, but conditions they have been exposed to.
The sensor data makes information about expiration more precise, Statler pointed out, than current forecasted “sell by” or “use by” dates.
“Those dates were set based on this theoretical perfect world where handling is done a certain way and we’ve increasingly found out that the handling of products is incredibly variable,” he said.
Ultimately, that kind of information could help a retailer reduce waste by ensuring that a product doesn’t spoil before it is put out for sale.
Future Applications in the Home
Wiliot is considering future use cases that could extend from the retailer to the consumer. Some smart appliances are already being built with Bluetooth functionality.
If those BLE radios in a refrigerator could energize and read tags on packaged food, that data could be managed with WiliBot for the benefit of consumers themselves.
One example: if an item in the fridge is set to expire, or even if a sensor detects spoilage, that data could then be sent to a homeowner via the app. But there are still issues to be worked out before the becomes a reality.
“We’re working towards it in a number of meaningful ways: one is getting the infrastructure in place to satisfy the enterprise use cases that provide a solid ROI,” said Statler. that, “When consumers will end up talking to their strawberries and tofu is a matter of conjecture but we know that there’s huge opportunities.”
In the meantime, the focus is to build out the solution for supply chain and retailer management.
Synergy Between AI and IoT
AI and IoT data can be mutually beneficial, Wiliot argues. AI works best with lots of data and the ambient IoT systems are generating millions of data point.
Wiliot is now using WiliBot as standard in in their pilots “because it basically allows us to show value really quickly,” said Statler.
Users of the system include those selling fresh foods as well as wine, medicine or managing luggage.
For that reason, Statler sees an “opportunity to make a difference to the cost of goods, to fight inflation, to improve health care, and to have safer better food and fight climate change.”