Uploaded August 2026 | Updated September 2026, 2 weeks ago
AI isn't the biggest limitation in supply chains anymore. The bigger problem is that AI still can't reliably sense what's happening in the physical world.
Most organisations are investing in smarter models while overlooking the data foundation those models depend on.
I'm joined by Doron Hazan, Director of Data Product and Artificial Intelligence at Wiliot, to examine what happens when AI moves beyond documents and dashboards into warehouses, distribution centres and logistics networks. His work sits at the intersection of artificial intelligence, ambient IoT and real-time operational data, helping organisations connect physical assets to AI-driven decision-making.
The timing couldn't be more relevant. Supply chain leaders are under pressure to improve service, reduce costs, manage disruption and justify AI investments that deliver measurable operational value. Yet many AI projects struggle because the underlying physical data is incomplete, delayed or disconnected from reality. If AI can't accurately perceive the environment it's operating in, better models alone won't solve the problem.
One idea particularly stayed with me throughout this discussion: today's AI has developed an incredibly capable "brain", but it still lacks a "nervous system". That analogy neatly explains why AI performs so well in digital environments yet often struggles in physical operations. We also explore why dashboards can become misleading when the underlying data breaks down, why supply chains remain far more chaotic than many AI practitioners expect, and why continuous sensing may become the missing layer between AI and genuinely autonomous operations.
If you're responsible for supply chain, procurement, logistics, manufacturing, operations, digital transformation or AI strategy, I think you'll find plenty here to challenge some common assumptions.
If you're dealing with these challenges in your own organisation, I'd love to hear how you're approaching them in the comments.
🌐 Podcast website: resilientsupplychainpodcast.com
🎧 Subscribe for weekly conversations on supply chain resilience, operational risk, AI, sustainability and decision-making.
Chapters:
00:00 Agents Need Senses
00:42 Podcast Intro and Guest
02:00 What Wiliot Does
02:54 Why Physical World Matters
04:31 Urgency of Supply Chain AI
05:36 When Reality Disagrees
06:55 Robots Without Context
08:15 Items as Live Data
09:44 Defining Physical AI
10:52 Sensing Value vs Overkill
13:12 Fix Data Before AI
14:12 Deploying Beyond Pilots
16:12 Dashboards vs Dock Reality
17:14 Chaos and Visibility Gaps
19:40 Humans in the Loop
21:37 Next Three to Five Years
22:38 Autonomy Risks and Upside
24:14 What Continuous Data Enables
26:08 How to Get Started
27:43 Key Takeaways for Leaders
28:33 Leadership and Churchill
30:01 Wrap Up and Contact
AI isn't the biggest limitation in supply chains anymore. The bigger problem is that AI still can't reliably sense what's happening in the physical world.
Most organisations are investing in smarter models while overlooking the data foundation those models depend on.
I'm joined by Doron Hazan, Director of Data Product and Artificial Intelligence at Wiliot, to examine what happens when AI moves beyond documents and dashboards into warehouses, distribution centres and logistics networks. His work sits at the intersection of artificial intelligence, ambient IoT and real-time operational data, helping organisations connect physical assets to AI-driven decision-making.
The timing couldn't be more relevant. Supply chain leaders are under pressure to improve service, reduce costs, manage disruption and justify AI investments that deliver measurable operational value. Yet many AI projects struggle because the underlying physical data is incomplete, delayed or disconnected from reality. If AI can't accurately perceive the environment it's operating in, better models alone won't solve the problem.
One idea particularly stayed with me throughout this discussion: today's AI has developed an incredibly capable "brain", but it still lacks a "nervous system". That analogy neatly explains why AI performs so well in digital environments yet often struggles in physical operations. We also explore why dashboards can become misleading when the underlying data breaks down, why supply chains remain far more chaotic than many AI practitioners expect, and why continuous sensing may become the missing layer between AI and genuinely autonomous operations.
If you're responsible for supply chain, procurement, logistics, manufacturing, operations, digital transformation or AI strategy, I think you'll find plenty here to challenge some common assumptions.
If you're dealing with these challenges in your own organisation, I'd love to hear how you're approaching them in the comments.
🌐 Podcast website: resilientsupplychainpodcast.com
🎧 Subscribe for weekly conversations on supply chain resilience, operational risk, AI, sustainability and decision-making.
Chapters:
00:00 Agents Need Senses
00:42 Podcast Intro and Guest
02:00 What Wiliot Does
02:54 Why Physical World Matters
04:31 Urgency of Supply Chain AI
05:36 When Reality Disagrees
06:55 Robots Without Context
08:15 Items as Live Data
09:44 Defining Physical AI
10:52 Sensing Value vs Overkill
13:12 Fix Data Before AI
14:12 Deploying Beyond Pilots
16:12 Dashboards vs Dock Reality
17:14 Chaos and Visibility Gaps
19:40 Humans in the Loop
21:37 Next Three to Five Years
22:38 Autonomy Risks and Upside
24:14 What Continuous Data Enables
26:08 How to Get Started
27:43 Key Takeaways for Leaders
28:33 Leadership and Churchill
30:01 Wrap Up and Contact
