To ensure that its data continues to improve as autonomous vehicles hit the market, Convoy could aim to equip these self driving trucks with IoT technology — perhaps the two are not mutually exclusive.
There are certainly barriers to the digital transformation of this historically offline industry. Hopefully the timing is right and the economic benefits of digitization will overcome this friction.
You make an important point about self driving trucks. While Convoy certainly needs to consider the long-term implications of autonomous vehicles, I believe the complexity of shipping will require humans to accompany trucks for the foreseeable future. As a simple example, who will ensure that cargo is correctly packed and the back of the truck is secure? 
As you alluded to, it is often very difficult for a software company to produce hardware. This is because the two product categories thrive under different organizational architectures (people, culture, tasks, structure).
However, that’s not to say it is impossible. There are many examples of companies, such as Adobe,  that have pulled it off. If Convoy were able to produce a simple IoT line of its own, then it would have an easier time standardizing the data it collects across carriers.
Matt, I agree that a more efficient logistics operation will reduce emissions. In fact Anheuser-Busch specifically cited environmental impact as one of its reasons for forming a long term partnership with Convoy! 
A couple questions have been posed in the comments to this article: Will the US factory survive long term, without government incentives? Can Foxconn keep labor-intensive activities in China while performing other activities in the US?
Foxconn may be able to apply research and development to modify its products and processes in accordance with the comparative advantage of the United States.  We often see this phenomenon as technologies are transferred from one country into another. For example, a Chinese company importing a manufacturing process that originated in the United States could apply R&D to adapt that processes in a labor-using fashion (e.g. more workers, fewer machines to produce the same output). Or, as the price of materials increases relative to the price of value-added inputs such as capital and labor, a company may apply R&D to ascend the value chain and use fewer materials.
Therefore, Foxconn may be able to succeed in the United States by applying R&D to innovate its products and processes in accordance with the comparative advantage of the United States. To do this it will need to conduct detailed pricing and productivity analysis of its factors of production in both the US and China.
 Atkinson, A. B., & Stiglitz, J. E. (1969). A New View of Technological Change. The Economic Journal , 79 (315), 573-578.
The notion of global distribution to narrow-interest audiences around the world is exciting. However, fear the likelihood of CJ E&M’s content to be found via recommendations will decline as streaming platforms increasingly favor their own content.  To move the needle with such small audiences around the world, CJ E&M may need to invest heavily in localizing its content to many countries. I feel more comfortable with the approach the company is taking in Vietnam, with a very aggressive plan to sweep a specific geographic market.
To the question of how Wal-Mart might adapt to climate change rather than mitigate, they might consider locating new stores in areas less likely to be affected by natural disasters and rising sea levels. Similarly, they may evaluate and influence the locations of warehouses and key logistic nodes. The risks related to location are particularly pronounced at JFK airport. 
Just as it uses it’s bargaining power to push suppliers to combat climate change, Wal-Mart can use that same influence to influence partners to adapt.
Your question on who should bear the costs of upgrade projects at JFK related to climate change and disaster mitigation reminded me of another key actor: the insurer. What role can the insurance industry play in preparing for climate change?
Perhaps insurers can adjust pricing to incentivize climate change preparedness. There is some evidence that works for homeowners — perhaps it can also work for commercial entities such as JFK. However, this will require an ability to collect an use relevant data to predict risk. 
M&M’s Connected Factory of the Future is quite impressive. I would love to learn more about the resources and analytics technology employed to derive actionable insights from their data.
The rise of IoT driven manufacturing creates a wealth of new data. But how difficult is it to make sense of that information? Some companies, like SAP, are creating software for supply chain control towers to help tackle this challenge.  It will be interesting to see how companies like M&M approach the issue.
As you pointed it, it will be difficult for C.H. Robinson to transform its fright brokerage offering into a competitive digital marketplace. Uber and other well funded startups such as Convoy, which recently raised $62M from investors,  are more competent with technology.
However, C.H. Robinson can add significant value to its digital marketplace by integrating it seamlessly with its TMC and 3PL solutions. For example, shippers that use pure play digital platforms such as Uber may have trouble integrating those platforms with their supply chain analytics centers. Whereas C.H. Robinson can ensure that it’s marketplace data is properly utilized by the shipper to improve its end-to-end supply chain.