Matt was using recipe 1, which had been changed so that there was no backside He in step 2. This is typically turned on in step 1 (He clamping.) The errors he got started with He backside, and cascaded into wafer transport problems, so there is reason to believe this was the cause. It was changed back.